System

Abstract

A system is provided.SOLUTION: The system includes a means for acquiring a camera video for identifying a person or an object, a means for analyzing camera video data and identifying the person or the object, a means for detecting abnormal behavior or an emergency situation, a means for automatically issuing an alarm on the basis of the detected abnormal behavior or emergency situation, and a means for cooperating with a related organization in real time.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The technology of the present disclosure relates to a system. [Background technology]

[0002] Patent document 1 discloses a persona chatbot control method performed by at least one processor, the method including the steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to a description of the chatbot character, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance in response to the user utterance. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-180282 Summary of the Invention [Problem to be solved by the invention]

[0004] Current surveillance camera systems have difficulty detecting abnormal behavior or emergencies in real time, limiting their ability to prevent crime and respond quickly to emergencies. They also lack support for appropriate resource allocation and efficient crime prevention activities. Furthermore, smooth collaboration with related agencies is difficult, sometimes resulting in delayed responses. To solve these problems, a high-performance and efficient surveillance camera system is needed. [Means for solving the problem]

[0005] To solve the above problems, the present invention provides the following means. The system includes a means for acquiring camera footage for identifying people or objects, and a means for analyzing the acquired camera footage data to identify people or objects. The analysis means can detect abnormal behavior or emergency situations. The system further includes a means for automatically issuing an alert based on the detected abnormal behavior or emergency situation. The system also includes a means for real-time collaboration with relevant organizations. Furthermore, big data analysis and predictive models can be used to predict and prevent abnormal behavior or emergency situations. When an alert is sent, detailed information including video clips, timestamps, and location information is provided, enabling relevant organizations to respond quickly. Furthermore, a notification of abnormal behavior or emergency situations is sent to a user terminal, providing a detailed explanation and a link to the video clip. The system further includes a means for compressing and transmitting camera footage. This significantly improves crime prevention and disaster response capabilities in cities, commercial districts, public transportation systems, and communities.

[0006] "Camera footage" refers to real-time visual data acquired by a surveillance camera.

[0007] "Analysis" is the process of identifying people and objects and detecting abnormal behavior and emergencies based on acquired camera video data.

[0008] "Abnormal behavior" refers to behavior that deviates from normal patterns of behavior, including falling, fighting, and stealing.

[0009] An "emergency" refers to a situation that requires a prompt response, such as a crime, fire, or accident.

[0010] "Alert" refers to an automated notification or alert that is issued when abnormal behavior or an emergency is detected.

[0011] "Related agencies" refers to organizations and groups that work together to respond to abnormal behavior or emergencies, such as police, fire departments, and security companies.

[0012] "Real time" refers to a time situation in which processing or response is carried out immediately without delay.

[0013] "Big data" refers to massive datasets that can be analyzed to extract useful information and patterns.

[0014] A "predictive model" is an algorithm or mathematical model used to forecast future events based on historical data and current conditions.

[0015] "Video clip" means a portion of video data within a particular time range that has been extracted to show abnormal behavior or an emergency situation.

[0016] A "timestamp" refers to information that indicates a specific point in time, such as the date and time of recorded data or an event.

[0017] "Location information" refers to information that indicates a specific place or location, and includes GPS data.

[0018] "Compression" is the process of reducing the size of data, often used to make it easier to transfer or store.

[0019] A "server" is a computer system that receives and analyzes camera video data, detects abnormal behavior, and manages notifications and alerts.

[0020] A "terminal" is a device that includes a camera and is responsible for acquiring camera images and sending them to a server. [Brief explanation of the drawings]

[0021] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2]1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0022] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0023] First, the terms used in the following description will be explained.

[0024] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0025] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0026] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0027] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0028] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0029] [First embodiment]

[0030] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0031] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0032] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0033] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0034] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0035] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0036] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0037] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0038] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0039] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0040] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0041] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0042] The system of the present invention is a high-performance surveillance camera system that utilizes AI, and can identify people and objects, detect abnormal behavior and emergency situations, and automatically issue an alarm. The following describes in detail an embodiment of the present invention.

[0043] System Configuration

[0044] The system includes the following main components:

[0045] 1. Terminal (surveillance camera)

[0046] 2. Server

[0047] 3. User device (smartphone or PC)

[0048] 4. Related agencies (police, fire department, etc.)

[0049] Acquiring and transmitting camera images

[0050] The devices (surveillance cameras) are placed in cities, commercial districts, public transportation facilities, and communities. These cameras capture images in real time and store them in a buffer frame by frame. The video data in the buffer is compressed and sent to a server at regular intervals.

[0051] Receiving and analyzing video data

[0052] The server receives the video data sent from the device. The received data is decompressed and input into an AI model. This model identifies people and objects, and detects abnormal behavior and emergencies. For example, it can detect situations where a person falls or a fight breaks out. The detected data is recorded in a database.

[0053] Detecting abnormal behavior and issuing alerts

[0054] When an abnormal behavior or emergency situation is detected, the server acquires the timestamp and location information and prepares it for the emergency notification system. The alert is generated as an alert message containing the video clip, timestamp, and location information, and is sent to the relevant authorities in the form of a push notification.

[0055] Cooperation with related organizations

[0056] The server waits for confirmation of the alert sent to the relevant authorities, and once confirmation is received, it provides the next important update (e.g., changes in the situation on the ground, additional video clips, etc.) in real time, allowing the relevant authorities to respond quickly and accurately.

[0057] User Notifications

[0058] The server sends notifications to specific user devices when an emergency or abnormal behavior occurs, including a detailed description of the event and a link to a video clip, allowing users to take immediate action.

[0059] Big Data Analysis and Prediction

[0060] The server performs big data analysis using accumulated video data and abnormal behavior detection logs. It trains predictive models to forecast crimes and emergencies. Based on the prediction results, it helps optimize resource allocation and plan countermeasures. This improves the efficiency of crime prevention and emergency response.

[0061] Specific examples

[0062] Example 1: Incident detection on public transport

[0063] The device (surveillance camera) continues to monitor the video near the ticket gate of the station. Among the people passing by, it detects one person holding a knife.

[0064] The server detects this anomalous behavior and sends a notification to the emergency notification system, which includes the video clip, a timestamp, and location information.

[0065] The relevant authorities (police) receive a notification and dispatch the nearest police officer to the scene. The server continues to monitor the footage from the camera and provides additional information to the relevant authorities.

[0066] The user (station staff) will receive a notification on their smartphone, and will be able to immediately grasp the situation and prepare a response.

[0067] Example 2: Suspicious behavior detection in a commercial area

[0068] The device (surveillance camera) monitors the video at the entrance of the shopping mall and detects the behavior of a person stealing goods from multiple stores.

[0069] The server detects this unusual behavior and sends an alert to the police, along with video clips and location information, for rapid response.

[0070] The relevant authorities (police) receive the notification and immediately dispatch the nearest police officers to the scene. The server updates the relevant authorities as the situation changes.

[0071] The user (store staff) will receive a notification on their smartphone, immediately check the security camera footage, and take appropriate action while waiting for the police to arrive.

[0072] This allows the AI-powered surveillance camera system to detect abnormal behavior in real time, strengthen cooperation with relevant agencies, and enable rapid response. It also utilizes big data analytics and predictive models to prevent crime and efficiently allocate resources.

[0073] The processing flow will be explained below.

[0074] Step 1:

[0075] The device (surveillance camera) captures real-time video of the specified area. The video is divided into frames and temporarily stored in a buffer.

[0076] Step 2:

[0077] The terminal (surveillance camera) compresses the video data in the buffer at regular intervals (e.g., every second) and sends it to a server via the Internet.

[0078] Step 3:

[0079] The server receives the video data sent from the device, decompresses it, and immediately inputs it into the AI ​​model.

[0080] Step 4:

[0081] The server analyzes the transmitted video data and identifies people and objects using an AI model that is optimized based on pre-trained data.

[0082] Step 5:

[0083] Based on the analysis results, the server tracks the movements of people and objects in the video and uses an abnormal behavior detection algorithm to detect abnormal behavior such as falls, fights, or the possession of knives.

[0084] Step 6:

[0085] When the server detects abnormal or emergency activity, it clips the frame of data and flags it as abnormal activity. The video clip, timestamp, and location information are saved.

[0086] Step 7:

[0087] If any unusual behavior or emergency is flagged, the server immediately generates an alert message, which includes details of the incident, video clips, timestamps, and location information.

[0088] Step 8:

[0089] The server sends an alert to the relevant authorities (police and fire departments) in the form of a push notification. After sending the alert, it waits for confirmation of receipt of the alert, and once confirmation is received, it proceeds to the next step.

[0090] Step 9:

[0091] The relevant authorities (police and fire departments) receive an alert notification and begin preparing to respond at the scene, dispatching the nearest police officers and firefighters to the scene.

[0092] Step 10:

[0093] The server monitors the response status of relevant agencies and continues to transmit video data in real time if the situation at the scene changes, providing the latest information.

[0094] Step 11:

[0095] The server sends notifications of emergencies or unusual behavior to specific user devices (e.g., commercial facility staff, security personnel), including detailed descriptions and links to video clips.

[0096] Step 12:

[0097] Users (such as commercial facility staff) receive notifications on their smartphones or PCs, immediately check the situation, and take action to ensure the safety of their surroundings and prevent damage, if necessary.

[0098] Step 13:

[0099] The server stores the accumulated video data and abnormal behavior detection logs, and uses them for big data analysis, using predictive models to forecast future crimes and emergencies.

[0100] Step 14:

[0101] The server generates detailed reports based on the analysis results to help optimize resource allocation and plan crime prevention activities, and the generated reports enable efficient countermeasures to be taken.

[0102] This enables the system to leverage AI for real-time monitoring and immediate response, significantly improving safety in cities, commercial districts, public transportation, and communities.

[0103] Example 1

[0104] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0105] Conventional surveillance camera systems have difficulty detecting abnormal behavior and emergencies in real time and providing prompt notification. Furthermore, due to insufficient collaboration with relevant agencies and insufficient predictive capabilities based on big data analysis, there were issues with the efficiency of crime prevention and emergency response. Furthermore, the notification functionality was limited, preventing users from taking immediate action.

[0106] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0107] In this invention, the server includes a device for acquiring camera footage, a device for compressing the acquired video data and sending it to the server, a device for decompressing the data received by the server and inputting it into an AI model to identify people or objects, a device for detecting abnormal behavior or emergencies, a device for recording information on the detected abnormal behavior or emergency in a database and generating an alert, a device for coordinating with relevant organizations in real time, and a device for sending notifications of emergency situations or abnormal behavior to specific user devices. This enables real-time detection of abnormal behavior or emergencies and rapid notification, strengthens cooperation with relevant organizations, and enables crime prevention and efficient resource allocation through predictions using big data analysis.

[0108] A "device for acquiring camera images" is a device that has the function of capturing images in real time in the environment in which it is installed and storing the image data.

[0109] The "device that compresses video data and transmits it to a server" is a device that has the function of compressing acquired video data and transmitting it to a server in an efficient format.

[0110] "Device that decompresses received data and inputs it into an AI model" refers to a device that has the function of decompressing compressed data received by the server and inputting it into an AI model.

[0111] A "device for identifying people or objects" is a device that has the function of identifying people or objects in video data using an AI model.

[0112] The "device for detecting abnormal behavior or emergency situations" is a device that has the function of detecting abnormal behavior or emergency situations that exceed a predetermined standard based on identified data.

[0113] An "alert generating device" is a device that has the function of generating and notifying an alert based on information on detected abnormal behavior or emergency situations.

[0114] The "device that cooperates with relevant organizations in real time" is a device that has the function of notifying relevant organizations of generated alerts in real time and ensuring cooperation.

[0115] A "device that transmits a notification to a specific user device" is a device that has a function of transmitting a notification to a specific user device when abnormal behavior or an emergency occurs.

[0116] A "device that uses big data analysis and predictive models" is a device that has the function of predicting future abnormal behavior or emergency situations using big data analysis and predictive models based on accumulated video data and abnormal behavior detection logs.

[0117] A "database recording device" is a device that has the function of storing information on detected abnormal behavior or emergency situations in a database.

[0118] MODE FOR CARRYING OUT THE INVENTION

[0119] The present invention relates to a high-performance surveillance camera system that utilizes AI. Hereinafter, an embodiment of the present invention will be described in detail.

[0120] System Configuration

[0121] The system mainly consists of the following components:

[0122] 1. Terminal (surveillance camera)

[0123] 2. Server

[0124] 3. User device (smartphone or PC)

[0125] 4. Related agencies (police, fire department, etc.)

[0126] Acquiring camera footage

[0127] The devices (surveillance cameras) are placed in cities, commercial districts, public transportation facilities, and communities. These cameras capture video in real time and store it in a buffer frame by frame. This stored video data is recompressed at regular intervals and sent to a server.

[0128] Video data compression and transmission

[0129] The device compresses the captured video data using a compression method such as H.264. The compressed data is then encrypted and securely sent to the server using the HTTPS protocol.

[0130] Receiving and decompressing video data

[0131] The server receives the compressed data sent from the device, stores it temporarily, and then decompresses it using decompression software such as FFmpeg to restore the original video.

[0132] Identification and detection with AI models

[0133] The server uses a deep learning framework (e.g., TensorFlow or PyTorch) to input the decompressed video data into an AI model, which then identifies people and objects and detects abnormal behavior and emergencies.

[0134] Recording abnormal behavior and generating alerts

[0135] The server records the detected abnormal behavior and emergency information in a database (e.g., MySQL, PostgreSQL), and simultaneously generates an alert message and sends it to the emergency notification system.

[0136] Sending and acknowledging alerts

[0137] The server sends the generated alert message to the relevant organizations in the form of a push notification. After sending, it waits for an ACK (acknowledgment) and, once it receives the ACK, provides additional information in real time.

[0138] User Notifications

[0139] The server notifies specific user devices of emergencies or abnormal behavior, including a detailed description of the event and a link to a video clip, allowing users to take immediate action.

[0140] Big Data Analysis and Prediction

[0141] The server uses the accumulated video data and abnormal behavior detection logs to perform big data analysis and predictive models. It processes the data using big data processing tools such as Hadoop and Spark to train predictive models. This makes it possible to predict crimes and emergencies, and helps with optimal resource allocation and countermeasure planning.

[0142] Specific examples

[0143] Example 1: Incident detection on public transport

[0144] The device (surveillance camera) continues to monitor the video near the ticket gate of the station. Among the people passing by, it detects one person holding a knife.

[0145] The server detects this unusual behavior and sends a notification to the emergency notification system, which includes a video clip, a timestamp, and location information.

[0146] The relevant authorities (police) are notified and dispatch the nearest police officers to the scene. The server continues to monitor the footage from the camera and provides additional information to the relevant authorities.

[0147] The user (station staff) will receive a notification on their smartphone, and will be able to immediately grasp the situation and prepare a response.

[0148] Prompt Sentence Examples

[0149] "Please describe an AI model that instantly detects a person with a knife at a station ticket gate and sends a notification to the appropriate authorities via an emergency notification system."

[0150] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0151] Step 1:

[0152] The terminals (surveillance cameras) capture images in real time in the areas where they are installed.

[0153] Input: Video data from a surveillance camera sensor

[0154] Data processing: Capture video frame by frame

[0155] Output: Video data divided into frames is stored in a buffer

[0156] Specific operation: The surveillance camera captures 30 frames of video per second and stores them in buffer memory.

[0157] Step 2:

[0158] The terminal compresses the video data in the buffer at regular intervals.

[0159] Input: Buffered raw video data

[0160] Data processing: Compresses video using compression algorithms such as H.264

[0161] Output: Compressed video data

[0162] Specific operation: Every 10 seconds, the device compresses the video data stored in the buffer and stores it in the output buffer.

[0163] Step 3:

[0164] The terminal transmits the compressed video data to the server.

[0165] Input: Compressed video data

[0166] Data processing: Encrypting data and preparing it for transmission using the HTTPS protocol

[0167] Output: Secure compressed video data sent to the server

[0168] Specific operation: The terminal encrypts the compressed video data using HTTPS and sends it to the server.

[0169] Step 4:

[0170] The server receives the compressed data sent from the terminal.

[0171] Input: Compressed and encrypted video data

[0172] Data processing: Decrypt the received data and store it in temporary storage

[0173] Output: Compressed data saved to storage

[0174] Specific operation: The server receives data sent via HTTPS, decrypts it, and stores it in storage.

[0175] Step 5:

[0176] The server decompresses the received data.

[0177] Input: Compressed video data

[0178] Data processing: Decompress the compressed data using decompression software such as FFmpeg.

[0179] Output: Original video data

[0180] Specific operation: The server uses FFmpeg to decompress the compressed data and return it to the original video frames.

[0181] Step 6:

[0182] The server inputs the decompressed video data into the AI ​​model.

[0183] Input: Decompressed video data

[0184] Data processing: Analyze video using deep learning frameworks such as TensorFlow and PyTorch

[0185] Output: Recognition results (identification of people and objects, detection of abnormal behavior)

[0186] How it works: The AI ​​model inputs video data frame by frame to detect abnormal behavior or emergencies.

[0187] Step 7:

[0188] The server records any abnormal behavior or emergency situations and generates alerts.

[0189] Input: Recognition results by AI model

[0190] Data processing: Recording abnormal behavior and emergency information in a database and generating alerts

[0191] Output: Data recorded in the database, generated alert messages

[0192] Specific operation: The server records abnormal behavior information in MySQL or PostgreSQL and generates an alert.

[0193] Step 8:

[0194] The server sends an alert message to the appropriate authority and waits for an acknowledgement.

[0195] Input: The generated alert message

[0196] Data processing: Generate and send a push notification containing the alert message

[0197] Output: Notification sent to the relevant authorities, acknowledgement (ACK)

[0198] Specific operation: The server sends a notification using REST API or WebSocket and waits for confirmation of receipt.

[0199] Step 9:

[0200] The server sends a notification to the specific user terminal.

[0201] Input: Information about abnormal behavior or emergency situations

[0202] Data processing: Generate notification messages and send them to the user terminal

[0203] Output: Notification sent to the user's device

[0204] Specific operation: The server uses a push notification service (e.g., Firebase Cloud Messaging) to send a notification to the user device.

[0205] Step 10:

[0206] The server performs big data analysis and predictions.

[0207] Input: Accumulated video data and abnormal behavior detection logs

[0208] Data processing: Analyze data and train predictive models using Hadoop or Spark

[0209] Output: Prediction results, optimal resource allocation proposals

[0210] Specific operation: The server processes the data using big data tools and trains a predictive model to predict future abnormal behavior.

[0211] (Application example 1)

[0212] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0213] In modern society, surveillance systems are important for ensuring public safety. However, current surveillance systems lack the ability to detect abnormal behavior and emergencies in real time, making it difficult for security guards and relevant agencies to respond immediately. Furthermore, they lack the functionality to quickly detect abnormalities on-site using mobile devices and respond appropriately, resulting in a decrease in the efficiency of security services.

[0214] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0215] In this invention, the server includes means for acquiring camera footage for identifying people or objects, means for analyzing camera footage data and identifying people or objects, means for detecting abnormal behavior or emergencies, means for automatically issuing an alarm based on the detected abnormal behavior or emergency, means for coordinating with relevant organizations in real time, means for displaying an alarm on a display device mounted on a mobile terminal, means for transmitting video data from the mobile terminal to the server, and means for transmitting analyzed data and alarm information to relevant organizations. This enables security guards to immediately recognize abnormal behavior on-site and respond quickly, strengthening cooperation with relevant organizations and ensuring public safety.

[0216] "Means for acquiring camera footage to identify people or objects" refers to the function of collecting video data using surveillance cameras.

[0217] "Means for analyzing camera video data and identifying people or objects" refers to algorithms or software for analyzing collected video data and identifying people or objects.

[0218] "Means for detecting abnormal behavior or emergency situations" refers to the ability to use AI or machine learning models to detect behavior that deviates from normal behavior or emergency situations in real time.

[0219] "Means for automatically issuing an alert based on detected abnormal behavior or emergency situations" refers to a function for automatically issuing an alert when abnormal behavior or emergency situations are detected.

[0220] "Means of coordinating with relevant agencies in real time" refers to a system for sharing information in real time with relevant agencies such as the police and fire department and taking joint action.

[0221] "Means for displaying an alert on a display device installed in a mobile terminal" refers to a function for displaying alerts or notification messages in real time on a display device such as a smartphone or smart glasses.

[0222] "Means for transmitting video data from a mobile terminal to a server" refers to a function for uploading video data from a mobile terminal to a server via Internet communication.

[0223] "Means for transmitting analyzed data and warning information to relevant organizations" refers to a system for instantly transmitting AI analysis results and warning notifications to designated relevant organizations.

[0224] MODE FOR CARRYING OUT THE INVENTION

[0225] The present invention provides a security service using a high-performance surveillance camera system and mobile devices that utilize AI. An embodiment of this system is described in detail below.

[0226] System Configuration

[0227] The system consists of the following main components:

[0228] 1. Surveillance camera (terminal)

[0229] 2. Mobile devices (e.g., smart glasses)

[0230] 3. Server

[0231] 4. User devices (smartphones and PCs)

[0232] 5. Related agencies (police, fire department, etc.)

[0233] Acquiring and transmitting camera images

[0234] Surveillance cameras: These cameras are installed in cities, commercial districts, and public facilities and have the ability to capture video in real time. These videos are compressed at regular intervals and sent to a server.

[0235] Receiving and analyzing video data

[0236] Server: Receives and decompresses video data sent from surveillance cameras. The data is then input into an AI model to detect abnormal behavior and emergencies. Examples of AI models used include YOLO and SSD.

[0237] Detection of abnormal behavior: When the server detects abnormal behavior or an emergency, it captures the timestamp and location information and records it in a database.

[0238] Alarm display on mobile device

[0239] Smart glasses: Alerts from the server and notifications about abnormal behavior are displayed on the mobile device in real time to warn the user.

[0240] Video data transmission: On-site video data is uploaded from the smart glasses or smartphone to a server for further analysis.

[0241] Cooperation with related organizations

[0242] Alert notification: The server generates an alert message based on the detected abnormal behavior and sends the analyzed data to the relevant authorities. The notification includes the video clip, timestamp, and location information.

[0243] Big Data Analysis and Prediction

[0244] Big data analysis: The server performs big data analysis using the accumulated video data and abnormal behavior logs, which makes it possible to predict and prevent crimes and emergencies.

[0245] Specific prompt examples

[0246] 1. Example 1: A person is in possession of a knife at a train station.

[0247] 2. Example 2: Theft is detected in a shopping mall.

[0248] Hardware and software used

[0249] Hardware:

[0250] Surveillance cameras: high-resolution cameras (e.g. CCTV cameras)

[0251] Mobile devices: smart glasses (e.g., Google Glass, Vuzix Blade), smartphones

[0252] software:

[0253] AI model: YOLO, SSD

[0254] Server: AWS, Google Cloud

[0255] Notification systems: Firebase, Twilio

[0256] Program processing

[0257] 1. The device (surveillance camera) captures video in real time and sends it to the server at regular intervals.

[0258] 2. The server receives the video data, analyzes it using an AI model, and detects abnormal behavior or emergencies.

[0259] 3. The server generates an alert based on the detected abnormal behavior or emergency situation and displays it on the mobile device (smart glasses, smartphone).

[0260] 4. Send the alert and detailed data to relevant authorities (police, fire department, etc.).

[0261] This will enable security guards to immediately recognize any abnormal behavior on-site and respond quickly, strengthening cooperation with relevant agencies and ensuring public safety.

[0262] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0263] Step 1:

[0264] The terminal (surveillance camera) captures video in real time and temporarily stores the video data in a buffer. At regular intervals, this video data is compressed and sent to the server. The input here is the raw video data captured by the camera, and the output is compressed video data.

[0265] Step 2:

[0266] The server receives the compressed video data sent from the terminal and decompresses it. The decompressed video data is ready for analysis frame by frame. The input here is the compressed video data and the output is the decompressed frame data.

[0267] Step 3:

[0268] The server inputs the decompressed video data into a generative AI model (e.g., YOLO or SSD) to identify people or objects. This identification data is then used to detect abnormal behavior or emergencies. Here, the input is the decompressed frame data, and the output is a list of identified objects.

[0269] Step 4:

[0270] The server analyzes the identified object list to detect anomalous behavior or emergencies. If an anomaly is detected, its timestamp and location information are recorded in a database. The input here is the list of identified objects, and the output is the detection of anomalous behavior or emergencies.

[0271] Step 5:

[0272] The server automatically issues an alert based on the detected abnormal behavior or emergency situation. This alert is sent to the display device of the mobile device (smart glasses or smartphone) and displays a warning to the user. The input here is the abnormal behavior detection result, and the output is the alert message.

[0273] Step 6:

[0274] A mobile device (smart glasses or smartphone) receives an alert message from the server and displays it to the user in real time. The user can then check the details of the alert and take necessary measures. The input here is the alert message, and the output is a notification display to the user.

[0275] Step 7:

[0276] The server sends the alert and detailed data (video clip, timestamp, location information) to the relevant authorities (police, fire department, etc.). The relevant authorities then take prompt action based on this information. The input here is the alert message and detailed data, and the output is a notification to the relevant authorities.

[0277] Step 8:

[0278] The server performs big data analysis and trains a predictive model based on the accumulated video data and abnormal behavior logs. This enables future predictions of crimes and emergencies and allows preventative measures to be planned. The input here is past video data and abnormal behavior logs, and the output is the results of the predictive model.

[0279] By taking the above steps, security guards can immediately recognize any abnormal behavior on-site and respond quickly, strengthening cooperation with relevant agencies and ensuring public safety.

[0280] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0281] The system of the present invention combines an emotion engine with a high-performance surveillance camera system that utilizes AI, and is capable of identifying people and objects, detecting abnormal behavior and emergency situations, automatically issuing alarms, and recognizing the user's emotions and responding based on those emotions. The following describes in detail the embodiments of the present invention.

[0282] System Configuration

[0283] The system includes the following main components:

[0284] 1. Terminal (surveillance camera)

[0285] 2. Server

[0286] 3. Emotion Engine

[0287] 4. User devices (smartphones and PCs)

[0288] 5. Related agencies (police, fire department, etc.)

[0289] Acquiring and transmitting camera images

[0290] The device (surveillance camera) captures real-time video of a specified area. The video is divided into frames and temporarily stored in a buffer. The video data in the buffer is compressed at regular intervals and sent to the server.

[0291] Receiving and analyzing video data

[0292] The server receives the video data sent from the device. The received data is decompressed and immediately input into the AI ​​model. This model identifies people and objects, and detects abnormal behavior and emergencies. The detected data is recorded in a database.

[0293] Detecting abnormal behavior and issuing alerts

[0294] The server captures and flags any abnormal behavior or emergency situations, capturing their timestamp and location information. The video clip, timestamp, and location information are saved and generated as an alert message, which is then sent to the appropriate authorities in the form of a push notification.

[0295] Introducing the Emotion Engine

[0296] The emotion engine uses the captured video data and abnormal behavior analysis data to recognize emotions from the user's facial expressions and voice. This emotion data is used to evaluate abnormal behavior and optimize responses.

[0297] Optimizing responses based on perceived emotions

[0298] The server uses the emotion data obtained from the emotion engine to optimize the content of the alert and emergency response. For example, if a user is in a panic state, the server sends an alert that increases the level of urgency to the relevant authorities.

[0299] User Notifications

[0300] User devices (smartphones and PCs) receive notifications when abnormal behavior or emergencies occur, including a detailed description of the event, a link to a video clip, and appropriate countermeasures based on the user's feelings.

[0301] Cooperation with related organizations

[0302] The server waits for confirmation of the alert sent to the relevant authorities, and once confirmation is received, it provides the next important update (e.g., changes in the situation on the ground, additional video clips, etc.) in real time, allowing the relevant authorities to respond quickly and accurately.

[0303] Big Data Analysis and Prediction

[0304] The server uses the accumulated video data and emotion data for big data analysis, training predictive models to forecast crimes and emergencies, and supporting optimal resource allocation and countermeasure planning based on the prediction results.

[0305] Specific examples

[0306] Example 1: Incident detection on public transport

[0307] The device (surveillance camera) continues to monitor the video near the ticket gate of the station. Among the people passing by, it detects one person holding a knife.

[0308] The server detects this abnormal behavior and uses an emotion engine to analyze the emotions of the person wielding the knife and those around them. If emotions of panic or fear are recognized, a high-level alert is generated and sent to the relevant authorities.

[0309] The relevant authorities (police) receive a notification and dispatch the nearest police officer to the scene. The server continues to monitor the footage from the camera and provides additional information to the relevant authorities.

[0310] The user (station staff) will receive a notification on their smartphone, and will be able to immediately grasp the situation and prepare a response.

[0311] Example 2: Suspicious behavior detection in a commercial area

[0312] The device (surveillance camera) monitors the video at the entrance of the shopping mall and detects the behavior of a person stealing goods from multiple stores.

[0313] The server detects this abnormal behavior and uses an emotion engine to analyze the emotions of the thief and those around them. If emotions of anxiety or confusion are recognized, an alert is generated suggesting appropriate responses and sent to the relevant authorities.

[0314] The relevant authorities (police) receive the notification and immediately dispatch the nearest police officers to the scene. The server updates the relevant authorities as the situation changes.

[0315] The user (store staff) will receive a notification on their smartphone, immediately check the security camera footage, and take appropriate action while waiting for the police to arrive.

[0316] In this way, by combining an emotion engine with an AI-powered surveillance camera system, abnormal behavior can be detected in real time, strengthening collaboration with users and relevant agencies, and enabling swift and accurate responses. By utilizing big data analysis and predictive models, crime prevention and efficient resource allocation become possible.

[0317] The processing flow will be explained below.

[0318] Step 1:

[0319] The device (surveillance camera) captures real-time video of the specified area. The video is divided into frames and temporarily stored in a buffer.

[0320] Step 2:

[0321] The terminal (surveillance camera) compresses the video data in the buffer at regular intervals (e.g., every second) and sends it to a server via the Internet.

[0322] Step 3:

[0323] The server receives the video data sent from the device, decompresses it, and immediately inputs it into the AI ​​model.

[0324] Step 4:

[0325] The server analyzes the transmitted video data and identifies people and objects using an AI model that is optimized based on pre-trained data.

[0326] Step 5:

[0327] Based on the analysis results, the server tracks the movements of people and objects in the video and uses an abnormal behavior detection algorithm to detect abnormal behavior such as falls, fights, or the possession of knives.

[0328] Step 6:

[0329] When the server detects abnormal or emergency activity, it clips the frame of data and flags it as abnormal activity. The video clip, timestamp, and location information are saved.

[0330] Step 7:

[0331] The server uses an emotion engine to analyze the user's facial expressions and voice based on the analysis results of the video data and abnormal behavior, and recognizes emotions.

[0332] Step 8:

[0333] The server then optimizes the content of the alert and emergency response based on the recognized emotion. For example, if a user is in a panic state, it will generate a high-urgency alert and send it to the relevant authorities.

[0334] Step 9:

[0335] The server sends an alert to the relevant authorities (police, fire department) in the form of a push notification. The alert message includes details of the incident, a video clip, a timestamp, and location information.

[0336] Step 10:

[0337] The relevant authorities (police and fire departments) receive an alert notification and begin preparing to respond to the incident, dispatching the nearest police officers and firefighters to the scene.

[0338] Step 11:

[0339] The server monitors the response status of relevant agencies and continues to transmit video data in real time if the situation at the scene changes, providing the latest information.

[0340] Step 12:

[0341] The server sends notifications of emergencies or abnormal behavior to specific user devices (e.g., commercial facility staff, security personnel), including detailed descriptions, links to video clips, and appropriate countermeasures based on the recognized emotions.

[0342] Step 13:

[0343] Users (such as commercial facility staff) receive notifications on their smartphones or PCs, immediately check the situation, and take action to ensure the safety of their surroundings and prevent damage, if necessary.

[0344] Step 14:

[0345] The server stores the accumulated video data and emotion data and uses it for big data analysis, using predictive models to forecast future crimes and emergencies.

[0346] Step 15:

[0347] The server generates detailed reports based on the analysis results to help optimize resource allocation and plan crime prevention activities, and the generated reports enable efficient countermeasures to be taken.

[0348] This enables the system to utilize AI for real-time monitoring and immediate response, and employs an emotion engine to take appropriate action in the event of abnormal behavior or an emergency, significantly improving safety in cities, commercial districts, public transportation, and communities.

[0349] Example 2

[0350] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0351] Current surveillance camera systems are limited to simply monitoring video footage and detecting abnormal behavior, and lack the ability to accurately assess users' emotions and situations and respond quickly and appropriately. Furthermore, their functionality for coordinating with relevant agencies and preventing abnormal behavior is limited. This leads to problems such as delayed initial responses to emergencies and difficulty allocating appropriate resources.

[0352] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0353] In this invention, the server includes a means for acquiring camera footage, a means for analyzing the acquired camera footage in real time and identifying people or objects, a means for detecting abnormal behavior or emergencies, a means for automatically issuing an alert based on the detected abnormal behavior or emergency, a means for coordinating with relevant organizations in real time, and a means for recognizing emotions from the user's facial expressions and voice and optimizing responses. This integrates the process from detecting abnormal behavior to emergency response, enabling flexible responses based on the user's emotions. Furthermore, the combined use of big data analysis and predictive models enables more accurate prediction and prevention of abnormal behavior and emergencies, enabling optimal resource allocation and rapid response.

[0354] "Camera footage" refers to visual information obtained from a surveillance camera.

[0355] "Video data" refers to data that represents camera images in digital format.

[0356] "Real-time analysis" refers to processing and analyzing data as soon as it is received.

[0357] "Person or object identification" refers to analyzing surveillance camera footage and identifying people or objects present in the footage.

[0358] "Abnormal behavior" refers to behavior that deviates from general patterns of behavior and is considered dangerous or suspicious.

[0359] An "emergency" is a sudden situation or event that requires immediate action.

[0360] "Alert" refers to a warning or notification issued when abnormal behavior or an emergency is detected.

[0361] "Related agencies" refers to external organizations (e.g., police, fire department, etc.) that respond to abnormal behavior or emergencies.

[0362] "Facial expressions" refer to the movements and changes of facial muscles that express a person's emotions and psychological state.

[0363] "Audio" refers to data related to human voices and sounds.

[0364] "Emotion recognition" refers to identifying a user's emotions and mental state from their facial expressions and voice.

[0365] "Optimizing responses" refers to selecting and implementing the most appropriate and effective response measures based on emotional data and the situation.

[0366] "Big data analytics" refers to the process of analyzing large amounts of data and extracting meaningful information and patterns.

[0367] "Predictive model" refers to a mathematical or statistical model used to forecast future events or trends based on historical data.

[0368] The system of the present invention combines an emotion engine with a high-performance surveillance camera system that utilizes AI, and is capable of identifying people and objects, detecting abnormal behavior and emergency situations, automatically issuing alarms, and recognizing the user's emotions and responding based on those emotions. The following are modes for implementing the present invention.

[0369] System Configuration

[0370] The system includes the following main components:

[0371] 1. Terminal (surveillance camera)

[0372] 2. Server

[0373] 3. Emotion Engine

[0374] 4. User devices (smartphones and PCs)

[0375] 5. Related agencies (police, fire department, etc.)

[0376] Acquiring and transmitting camera images

[0377] The device (surveillance camera) captures real-time video of a specified area. For example, this applies to surveillance cameras installed in surveillance systems for public transportation or commercial districts. The video is divided into frames and temporarily stored in a buffer. The video data is periodically compressed using a compression technology such as H.264 and sent to a server using a secure protocol such as HTTPS.

[0378] Receiving and analyzing video data

[0379] The server receives the compressed video data sent from the device. The received data is decompressed and input into a generative AI model (e.g., YOLOv4 or OpenPose). This AI model identifies people and objects and detects abnormal behavior and emergencies. For example, it can detect dangerous behavior such as someone carrying a knife on public transportation.

[0380] Detecting abnormal behavior and issuing alerts

[0381] The server then assigns a timestamp and location information to any abnormal behavior or emergency detected by the AI ​​model. Based on this, an alert message is generated and sent to the relevant authorities in the form of a push notification, along with a detailed description of the incident and a link to a video clip. For example, police and fire departments are notified immediately.

[0382] Introducing an emotion engine and optimizing responses

[0383] The emotion engine analyzes the user's facial expressions and voice from the acquired video data and abnormal behavior analysis data to recognize their emotions. The server can use this emotion data to optimize the content of warnings and emergency responses. For example, if a user is in a panic state, it can send a high-level alert to the relevant authorities.

[0384] User notification and collaboration with relevant organizations

[0385] User devices (smartphones and PCs) receive notifications when abnormal behavior or emergencies occur. The notifications include a detailed description of the event, a link to a video clip, and appropriate countermeasures based on emotion data, allowing users to respond quickly.

[0386] Big Data Analysis and Prediction

[0387] The server then performs big data analysis on the accumulated video and emotion data, which can then be used to train predictive models to forecast future abnormal behavior and emergencies. The prediction results can then be used to optimize resource allocation and support countermeasure planning.

[0388] Resulting System Behavior

[0389] This system analyzes real-time footage from surveillance cameras, detects abnormal behavior and emergencies, and immediately coordinates with relevant agencies to take appropriate action. Furthermore, it has the ability to analyze user emotions and respond appropriately based on those emotions, enabling rapid and effective response to emergencies. Big data analysis also improves prediction accuracy, enabling advance countermeasures and optimal resource allocation.

[0390] Examples of prompt statements

[0391] "Please explain in detail the process of the system that detects abnormal behavior at station ticket gates and also analyzes the emotions of those around you."

[0392] "Please explain a system that detects suspicious behavior in a shopping mall and performs sentiment analysis based on that behavior."

[0393] In this way, the system of the present invention enables real-time detection of abnormal behavior, prompt notification to users and relevant organizations, and optimal response based on emotional data.

[0394] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0395] Step 1:

[0396] The terminal (surveillance camera) acquires real-time video of the designated area.

[0397] Input: Video of the specified area

[0398] Processing: Capture video in frames (e.g. 30 frames per second) and save as frame data

[0399] Output: Frame data

[0400] Step 2:

[0401] The terminal (surveillance camera) stores the acquired frame data in a buffer, compresses it at regular intervals, and sends it to the server.

[0402] Input: Frame data

[0403] Processing: Frame data is temporarily stored in a buffer and compressed using a compression algorithm such as H.264

[0404] Output: Compressed video data is sent to the server

[0405] Step 3:

[0406] The server receives the compressed video data sent from the terminal and decompresses it.

[0407] Input: Compressed video data

[0408] Processing: Decompresses received data and restores it as frame data

[0409] Output: Decompressed frame data

[0410] Step 4:

[0411] The server inputs the decompressed frame data into a generative AI model to identify people or objects.

[0412] Input: Decompressed frame data

[0413] Processing: Input frame data into a generative AI model (e.g., YOLOv4, OpenPose, etc.) to identify people and objects.

[0414] Output: Data of identified people or objects (location, label, etc.)

[0415] Step 5:

[0416] The server analyzes the data of the identified person or object to detect abnormal behavior or emergency situations.

[0417] Input: Identified person or object data

[0418] Processing: Uses analytical algorithms to detect abnormal behavior and emergencies

[0419] Output: Data on detected abnormal behavior or emergency (e.g., timestamp, location, behavior type, etc.)

[0420] Step 6:

[0421] The server issues alerts based on detected abnormal behavior or emergency situations and sends push notifications to relevant authorities.

[0422] Input: Data on abnormal behavior and emergencies

[0423] Processing: Generate an alert message and send it to the relevant authorities in the form of a push notification.

[0424] Output: Alert message to relevant organizations

[0425] Step 7:

[0426] The emotion engine uses the acquired video data and abnormal behavior analysis data to recognize emotions from the user's facial expressions and voice.

[0427] Input: Video data, abnormal behavior analysis data

[0428] Processing: Apply emotion recognition algorithms to identify the user's emotions

[0429] Output: Recognized emotion data (e.g., panic, fear, anxiety, etc.)

[0430] Step 8:

[0431] The server uses the recognized emotion data to optimize alert content and emergency response.

[0432] Input: Emotion data

[0433] Processing: Adjust the content and urgency of the warning message and optimize the necessary response measures.

[0434] Output: Optimized warning messages and countermeasures

[0435] Step 9:

[0436] User devices (smartphones and PCs) receive notifications when abnormal behavior or emergencies occur.

[0437] Input: Optimized warning message

[0438] Action: Display a notification and inform the user of the situation

[0439] Output: A message to inform the user

[0440] Step 10:

[0441] The server waits for confirmation of receipt of alerts sent to relevant authorities and provides important updates in real time.

[0442] Input: Acknowledgments from relevant agencies and new video data

[0443] Processing: Get acknowledgements and real-time updates on important information like changing conditions on the ground and additional footage clips

[0444] Output: Update information for relevant organizations

[0445] Step 11:

[0446] The server uses the accumulated video data and emotion data for big data analysis.

[0447] Input: Accumulated video data, emotion data

[0448] Processing: Data is stored in a data warehouse and analyzed using data mining techniques

[0449] Output: Analysis results and training data for predictive models

[0450] Step 12:

[0451] The server uses predictive models to forecast crimes and emergencies and help plan countermeasures.

[0452] Input: Prediction model, analysis results

[0453] Processing: Applying predictive algorithms to predict future abnormal behavior and emergencies

[0454] Output: Prediction results and supporting information for action plans

[0455] (Application example 2)

[0456] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0457] Conventional surveillance systems have the ability to identify people and objects and detect abnormal behavior and emergencies, but they are unable to recognize user emotions and provide responses based on those emotions. Furthermore, responses after detecting abnormal behavior are limited, often resulting in failure to contact relevant authorities or propose appropriate countermeasures. To address these issues, the present invention aims to provide a system that adds emotion recognition functionality and optimizes responses after detecting abnormal behavior.

[0458] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for acquiring camera images for identifying people or objects, means for analyzing camera image data and identifying people or objects, means for detecting abnormal behavior or emergency situations, and means for recognizing user emotions and optimizing responses based on the emotions. This makes it possible to take appropriate action based not only on identification but also on emotion data when abnormal behavior or an emergency occurs.

[0459] "Person or object identification" refers to technology that analyzes camera video data to identify people or objects.

[0460] "Unusual behavior or emergency" means any behavior that deviates from normal patterns of behavior or any situation that poses an immediate danger to life or property.

[0461] "Automatic alerting means" means a technology that automatically issues an alert in accordance with pre-defined conditions based on detected abnormal behavior or emergency situations.

[0462] "Means for recognizing the user's emotions and optimizing the response based on that" refers to technology that analyzes the user's emotions using acquired video and audio data and provides the optimal response based on the results.

[0463] "Means of coordinating with relevant agencies in real time" refers to technology that allows for instant communication with relevant agencies such as the police and fire department, providing necessary information, and maintaining a cooperative relationship.

[0464] "Big data analytics and predictive models" are techniques for analyzing large amounts of data to derive specific patterns and trends, which can be used to predict abnormal behavior and emergencies.

[0465] "Means for providing video clips, timestamps, and location information" refers to technology that, when issuing an alert, adds detailed information such as specific video data, the time of occurrence, and location.

[0466] "Means for proposing appropriate countermeasures" refers to technology that presents optimal and effective countermeasures based on detected abnormal behavior and user emotional data.

[0467] System Overview

[0468] The present invention is a system that identifies people and objects, detects abnormal behavior and emergencies, and automatically issues an alarm. Furthermore, it recognizes the user's emotions and provides an optimal response based on those emotions, making the response after abnormal behavior is detected more effective.

[0469] System Components

[0470] The system includes the following major components:

[0471] 1. Surveillance cameras

[0472] 2. Server

[0473] 3. Emotion Recognition Engine

[0474] 4. User devices (smartphones and PCs)

[0475] 5. Related agencies (police, fire department, etc.)

[0476] System Operation

[0477] Acquiring and transmitting camera images

[0478] The surveillance camera captures real-time video of a specified area. The video is divided into frames and temporarily stored in a buffer. The video data in the buffer is compressed at regular intervals and sent to the server.

[0479] Receiving and analyzing video data

[0480] The server receives video data transmitted from the surveillance cameras. The received data is decompressed and immediately input into a generative AI model. This model identifies people and objects, and detects abnormal behavior and emergencies. The detected data is then recorded in a database.

[0481] Detecting abnormal behavior and issuing alerts

[0482] The server captures and flags any abnormal behavior or emergency situations, capturing their timestamp and location information. The video clip, timestamp, and location information are saved, and an alert message is generated and sent to the appropriate authorities in the form of a push notification.

[0483] Using the Emotion Recognition Engine

[0484] The server uses the acquired video data and abnormal behavior analysis data to recognize emotions from the user's facial expressions and voice using an emotion recognition engine. This emotion data is used to evaluate abnormal behavior and optimize responses.

[0485] Optimizing responses based on perceived emotions

[0486] The server uses the emotion data obtained from the emotion recognition engine to optimize the content of the alert and emergency response, for example, sending an alert to the relevant authorities that increases the level of urgency if the user is in a panic.

[0487] User notification and response

[0488] The user device receives a notification when abnormal behavior or an emergency occurs. The notification includes a detailed description of the event, a link to a video clip, and appropriate countermeasures based on the user's emotions. The notification arrives on the user's smartphone, allowing the user to immediately understand the situation and prepare a response.

[0489] Hardware and software used

[0490] Hardware:

[0491] Surveillance cameras (IP cameras, etc.)

[0492] Smartphone or PC

[0493] software:

[0494] OpenCV: Camera image acquisition and analysis

[0495] Keras: Loading and predicting emotion recognition models

[0496] requests: Server alert notification

[0497] Specific processing examples

[0498] Example 1: Suspicious behavior detection in a shopping mall

[0499] 1. A surveillance camera monitors footage at the entrance to a shopping mall and detects an individual stealing items from multiple stores.

[0500] 2. The server detects this abnormal behavior and uses an emotion recognition engine to analyze the emotions of the thief and those around them. If emotions of anxiety or confusion are recognized, an alert is generated suggesting appropriate responses and sent to the relevant authorities.

[0501] 3. The relevant authorities (police) receive the notification and immediately dispatch the nearest police officers to the scene. The server updates the relevant authorities with information as the situation changes.

[0502] 4. The user (store staff) receives a notification on their smartphone, immediately checks the security camera footage, and takes appropriate action while waiting for the police to arrive.

[0503] Prompt Sentence Examples

[0504] "Detect anomalous behavior in this camera footage and recognize the emotions of people around it. If anomalous behavior is detected, generate an alert message suggesting appropriate countermeasures based on the emotional data."

[0505] The above is an embodiment of the present invention. The present invention makes it possible to detect abnormal behavior in real time, strengthen cooperation with users and related organizations, and realize a prompt and accurate response.

[0506] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0507] Step 1:

[0508] The surveillance camera captures real-time video of a specified area. The input is the video data from the surveillance camera, and the output is the video frame stored in the buffer. This frame is temporarily stored for further processing.

[0509] Step 2:

[0510] The server compresses the video data in the buffer at regular intervals and sends it to the server. The input is the video frames stored in the buffer, and the output is the compressed video data. Data compression enables efficient use of network bandwidth.

[0511] Step 3:

[0512] The server receives the compressed video data and decompresses it. The input is the compressed video data, and the output is the decompressed video data. The decompressed data is immediately input into the generative AI model.

[0513] Step 4:

[0514] The server uses a generative AI model to identify people and objects. The input is the decompressed video data, and the output is the information of the identified people and objects. The generative AI model identifies people and objects based on video analysis.

[0515] Step 5:

[0516] The server further analyzes the information of the identified people and objects to detect abnormal behavior or emergency situations. The input is the information of the identified people and objects, and the output is a flag indicating abnormal behavior or emergency situation. An algorithm is used to detect patterns of abnormal behavior or emergency situations.

[0517] Step 6:

[0518] When an abnormal behavior or emergency is detected, the server acquires the timestamp and location information and sets a flag. The input is the flag for abnormal behavior detection or emergency detection, and the output is the timestamp and location information.

[0519] Step 7:

[0520] The server generates an alert message based on the video clip, timestamp, and location information, and sends it to the relevant organizations. The input is the timestamp, location information, and video clip, and the output is the generated alert message. It is sent to the relevant organizations in the form of a push notification.

[0521] Step 8:

[0522] The server uses an emotion recognition engine to analyze the user's facial expressions and voice based on the acquired video data and recognizes their emotions. The input is video data and the output is emotion data. The emotion recognition engine estimates the user's emotions from their facial expressions and voice.

[0523] Step 9:

[0524] The server optimizes the content of the alert and emergency response using the emotion data obtained from the emotion recognition engine. The input is emotion data, and the output is optimized alert content and emergency response measures. The urgency and content of the response are adjusted based on the emotion data.

[0525] Step 10:

[0526] The user device receives notifications when abnormal behavior or an emergency occurs. The input is the optimized alert content and emergency response measures, and the output is a notification displayed on the user device. The user receives the notification on their smartphone, allowing them to immediately understand the situation and prepare a response.

[0527] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0528] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0529] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0530] [Second embodiment]

[0531] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0532] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0533] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0534] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0535] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0536] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0537] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0538] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0539] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0540] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0541] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0542] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0543] The system of the present invention is a high-performance surveillance camera system that utilizes AI, and can identify people and objects, detect abnormal behavior and emergency situations, and automatically issue an alarm. The following describes in detail an embodiment of the present invention.

[0544] System Configuration

[0545] The system includes the following main components:

[0546] 1. Terminal (surveillance camera)

[0547] 2. Server

[0548] 3. User device (smartphone or PC)

[0549] 4. Related agencies (police, fire department, etc.)

[0550] Acquiring and transmitting camera images

[0551] The devices (surveillance cameras) are placed in cities, commercial districts, public transportation facilities, and communities. These cameras capture images in real time and store them in a buffer frame by frame. The video data in the buffer is compressed and sent to a server at regular intervals.

[0552] Receiving and analyzing video data

[0553] The server receives the video data sent from the device. The received data is decompressed and input into an AI model. This model identifies people and objects, and detects abnormal behavior and emergencies. For example, it can detect situations where a person falls or a fight breaks out. The detected data is recorded in a database.

[0554] Detecting abnormal behavior and issuing alerts

[0555] When an abnormal behavior or emergency situation is detected, the server acquires the timestamp and location information and prepares it for the emergency notification system. The alert is generated as an alert message containing the video clip, timestamp, and location information, and is sent to the relevant authorities in the form of a push notification.

[0556] Cooperation with related organizations

[0557] The server waits for confirmation of the alert sent to the relevant authorities, and once confirmation is received, it provides the next important update (e.g., changes in the situation on the ground, additional video clips, etc.) in real time, allowing the relevant authorities to respond quickly and accurately.

[0558] User Notifications

[0559] The server sends notifications to specific user devices when an emergency or abnormal behavior occurs, including a detailed description of the event and a link to a video clip, allowing users to take immediate action.

[0560] Big Data Analysis and Prediction

[0561] The server performs big data analysis using accumulated video data and abnormal behavior detection logs. It trains predictive models to forecast crimes and emergencies. Based on the prediction results, it helps optimize resource allocation and plan countermeasures. This improves the efficiency of crime prevention and emergency response.

[0562] Specific examples

[0563] Example 1: Incident detection on public transport

[0564] The device (surveillance camera) continues to monitor the video near the ticket gate of the station. Among the people passing by, it detects one person holding a knife.

[0565] The server detects this anomalous behavior and sends a notification to the emergency notification system, which includes the video clip, a timestamp, and location information.

[0566] The relevant authorities (police) receive a notification and dispatch the nearest police officer to the scene. The server continues to monitor the footage from the camera and provides additional information to the relevant authorities.

[0567] The user (station staff) will receive a notification on their smartphone, and will be able to immediately grasp the situation and prepare a response.

[0568] Example 2: Suspicious behavior detection in a commercial area

[0569] The device (surveillance camera) monitors the video at the entrance of the shopping mall and detects the behavior of a person stealing goods from multiple stores.

[0570] The server detects this unusual behavior and sends an alert to the police, along with video clips and location information, for rapid response.

[0571] The relevant authorities (police) receive the notification and immediately dispatch the nearest police officers to the scene. The server updates the relevant authorities as the situation changes.

[0572] The user (store staff) will receive a notification on their smartphone, immediately check the security camera footage, and take appropriate action while waiting for the police to arrive.

[0573] This allows the AI-powered surveillance camera system to detect abnormal behavior in real time, strengthen cooperation with relevant agencies, and enable rapid response. It also utilizes big data analytics and predictive models to prevent crime and efficiently allocate resources.

[0574] The processing flow will be explained below.

[0575] Step 1:

[0576] The device (surveillance camera) captures real-time video of the specified area. The video is divided into frames and temporarily stored in a buffer.

[0577] Step 2:

[0578] The terminal (surveillance camera) compresses the video data in the buffer at regular intervals (e.g., every second) and sends it to a server via the Internet.

[0579] Step 3:

[0580] The server receives the video data sent from the device, decompresses it, and immediately inputs it into the AI ​​model.

[0581] Step 4:

[0582] The server analyzes the transmitted video data and identifies people and objects using an AI model that is optimized based on pre-trained data.

[0583] Step 5:

[0584] Based on the analysis results, the server tracks the movements of people and objects in the video and uses an abnormal behavior detection algorithm to detect abnormal behavior such as falls, fights, or the possession of knives.

[0585] Step 6:

[0586] When the server detects abnormal or emergency activity, it clips the frame of data and flags it as abnormal activity. The video clip, timestamp, and location information are saved.

[0587] Step 7:

[0588] If any unusual behavior or emergency is flagged, the server immediately generates an alert message, which includes details of the incident, video clips, timestamps, and location information.

[0589] Step 8:

[0590] The server sends an alert to the relevant authorities (police and fire departments) in the form of a push notification. After sending the alert, it waits for confirmation of receipt of the alert, and once confirmation is received, it proceeds to the next step.

[0591] Step 9:

[0592] The relevant authorities (police and fire departments) receive an alert notification and begin preparing to respond at the scene, dispatching the nearest police officers and firefighters to the scene.

[0593] Step 10:

[0594] The server monitors the response status of relevant agencies and continues to transmit video data in real time if the situation at the scene changes, providing the latest information.

[0595] Step 11:

[0596] The server sends notifications of emergencies or unusual behavior to specific user devices (e.g., commercial facility staff, security personnel), including detailed descriptions and links to video clips.

[0597] Step 12:

[0598] Users (such as commercial facility staff) receive notifications on their smartphones or PCs, immediately check the situation, and take action to ensure the safety of their surroundings and prevent damage, if necessary.

[0599] Step 13:

[0600] The server stores the accumulated video data and abnormal behavior detection logs, and uses them for big data analysis, using predictive models to forecast future crimes and emergencies.

[0601] Step 14:

[0602] The server generates detailed reports based on the analysis results to help optimize resource allocation and plan crime prevention activities, and the generated reports enable efficient countermeasures to be taken.

[0603] This enables the system to leverage AI for real-time monitoring and immediate response, significantly improving safety in cities, commercial districts, public transportation, and communities.

[0604] Example 1

[0605] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0606] Conventional surveillance camera systems have difficulty detecting abnormal behavior and emergencies in real time and providing prompt notification. Furthermore, due to insufficient collaboration with relevant agencies and insufficient predictive capabilities based on big data analysis, there were issues with the efficiency of crime prevention and emergency response. Furthermore, the notification functionality was limited, preventing users from taking immediate action.

[0607] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0608] In this invention, the server includes a device for acquiring camera footage, a device for compressing the acquired video data and sending it to the server, a device for decompressing the data received by the server and inputting it into an AI model to identify people or objects, a device for detecting abnormal behavior or emergencies, a device for recording information on the detected abnormal behavior or emergency in a database and generating an alert, a device for coordinating with relevant organizations in real time, and a device for sending notifications of emergency situations or abnormal behavior to specific user devices. This enables real-time detection of abnormal behavior or emergencies and rapid notification, strengthens cooperation with relevant organizations, and enables crime prevention and efficient resource allocation through predictions using big data analysis.

[0609] A "device for acquiring camera images" is a device that has the function of capturing images in real time in the environment in which it is installed and storing the image data.

[0610] The "device that compresses video data and transmits it to a server" is a device that has the function of compressing acquired video data and transmitting it to a server in an efficient format.

[0611] "Device that decompresses received data and inputs it into an AI model" refers to a device that has the function of decompressing compressed data received by the server and inputting it into an AI model.

[0612] A "device for identifying people or objects" is a device that has the function of identifying people or objects in video data using an AI model.

[0613] The "device for detecting abnormal behavior or emergency situations" is a device that has the function of detecting abnormal behavior or emergency situations that exceed a predetermined standard based on identified data.

[0614] An "alert generating device" is a device that has the function of generating and notifying an alert based on information on detected abnormal behavior or emergency situations.

[0615] The "device that cooperates with relevant organizations in real time" is a device that has the function of notifying relevant organizations of generated alerts in real time and ensuring cooperation.

[0616] A "device that transmits a notification to a specific user device" is a device that has a function of transmitting a notification to a specific user device when abnormal behavior or an emergency occurs.

[0617] A "device that uses big data analysis and predictive models" is a device that has the function of predicting future abnormal behavior or emergency situations using big data analysis and predictive models based on accumulated video data and abnormal behavior detection logs.

[0618] A "database recording device" is a device that has the function of storing information on detected abnormal behavior or emergency situations in a database.

[0619] MODE FOR CARRYING OUT THE INVENTION

[0620] The present invention relates to a high-performance surveillance camera system that utilizes AI. Hereinafter, an embodiment of the present invention will be described in detail.

[0621] System Configuration

[0622] The system mainly consists of the following components:

[0623] 1. Terminal (surveillance camera)

[0624] 2. Server

[0625] 3. User device (smartphone or PC)

[0626] 4. Related agencies (police, fire department, etc.)

[0627] Acquiring camera footage

[0628] The devices (surveillance cameras) are placed in cities, commercial districts, public transportation facilities, and communities. These cameras capture video in real time and store it in a buffer frame by frame. This stored video data is recompressed at regular intervals and sent to a server.

[0629] Video data compression and transmission

[0630] The device compresses the captured video data using a compression method such as H.264. The compressed data is then encrypted and securely sent to the server using the HTTPS protocol.

[0631] Receiving and decompressing video data

[0632] The server receives the compressed data sent from the device, stores it temporarily, and then decompresses it using decompression software such as FFmpeg to restore the original video.

[0633] Identification and detection with AI models

[0634] The server uses a deep learning framework (e.g., TensorFlow or PyTorch) to input the decompressed video data into an AI model, which then identifies people and objects and detects abnormal behavior and emergencies.

[0635] Recording abnormal behavior and generating alerts

[0636] The server records the detected abnormal behavior and emergency information in a database (e.g., MySQL, PostgreSQL), and simultaneously generates an alert message and sends it to the emergency notification system.

[0637] Sending and acknowledging alerts

[0638] The server sends the generated alert message to the relevant organizations in the form of a push notification. After sending, it waits for an ACK (acknowledgment) and, once it receives the ACK, provides additional information in real time.

[0639] User Notifications

[0640] The server notifies specific user devices of emergencies or abnormal behavior, including a detailed description of the event and a link to a video clip, allowing users to take immediate action.

[0641] Big Data Analysis and Prediction

[0642] The server uses the accumulated video data and abnormal behavior detection logs to perform big data analysis and predictive models. It processes the data using big data processing tools such as Hadoop and Spark to train predictive models. This makes it possible to predict crimes and emergencies, and helps with optimal resource allocation and countermeasure planning.

[0643] Specific examples

[0644] Example 1: Incident detection on public transport

[0645] The device (surveillance camera) continues to monitor the video near the ticket gate of the station. Among the people passing by, it detects one person holding a knife.

[0646] The server detects this unusual behavior and sends a notification to the emergency notification system, which includes a video clip, a timestamp, and location information.

[0647] The relevant authorities (police) are notified and dispatch the nearest police officers to the scene. The server continues to monitor the footage from the camera and provides additional information to the relevant authorities.

[0648] The user (station staff) will receive a notification on their smartphone, and will be able to immediately grasp the situation and prepare a response.

[0649] Prompt Sentence Examples

[0650] "Please describe an AI model that instantly detects a person with a knife at a station ticket gate and sends a notification to the appropriate authorities via an emergency notification system."

[0651] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0652] Step 1:

[0653] The terminals (surveillance cameras) capture images in real time in the areas where they are installed.

[0654] Input: Video data from a surveillance camera sensor

[0655] Data processing: Capture video frame by frame

[0656] Output: Video data divided into frames is stored in a buffer

[0657] Specific operation: The surveillance camera captures 30 frames of video per second and stores them in buffer memory.

[0658] Step 2:

[0659] The terminal compresses the video data in the buffer at regular intervals.

[0660] Input: Buffered raw video data

[0661] Data processing: Compresses video using compression algorithms such as H.264

[0662] Output: Compressed video data

[0663] Specific operation: Every 10 seconds, the device compresses the video data stored in the buffer and stores it in the output buffer.

[0664] Step 3:

[0665] The terminal transmits the compressed video data to the server.

[0666] Input: Compressed video data

[0667] Data processing: Encrypting data and preparing it for transmission using the HTTPS protocol

[0668] Output: Secure compressed video data sent to the server

[0669] Specific operation: The terminal encrypts the compressed video data using HTTPS and sends it to the server.

[0670] Step 4:

[0671] The server receives the compressed data sent from the terminal.

[0672] Input: Compressed and encrypted video data

[0673] Data processing: Decrypt the received data and store it in temporary storage

[0674] Output: Compressed data saved to storage

[0675] Specific operation: The server receives data sent via HTTPS, decrypts it, and stores it in storage.

[0676] Step 5:

[0677] The server decompresses the received data.

[0678] Input: Compressed video data

[0679] Data processing: Decompress the compressed data using decompression software such as FFmpeg.

[0680] Output: Original video data

[0681] Specific operation: The server uses FFmpeg to decompress the compressed data and return it to the original video frames.

[0682] Step 6:

[0683] The server inputs the decompressed video data into the AI ​​model.

[0684] Input: Decompressed video data

[0685] Data processing: Analyze video using deep learning frameworks such as TensorFlow and PyTorch

[0686] Output: Recognition results (identification of people and objects, detection of abnormal behavior)

[0687] How it works: The AI ​​model inputs video data frame by frame to detect abnormal behavior or emergencies.

[0688] Step 7:

[0689] The server records any abnormal behavior or emergency situations and generates alerts.

[0690] Input: Recognition results by AI model

[0691] Data processing: Recording abnormal behavior and emergency information in a database and generating alerts

[0692] Output: Data recorded in the database, generated alert messages

[0693] Specific operation: The server records abnormal behavior information in MySQL or PostgreSQL and generates an alert.

[0694] Step 8:

[0695] The server sends an alert message to the appropriate authority and waits for an acknowledgement.

[0696] Input: The generated alert message

[0697] Data processing: Generate and send a push notification containing the alert message

[0698] Output: Notification sent to the relevant authorities, acknowledgement (ACK)

[0699] Specific operation: The server sends a notification using REST API or WebSocket and waits for confirmation of receipt.

[0700] Step 9:

[0701] The server sends a notification to the specific user terminal.

[0702] Input: Information about abnormal behavior or emergency situations

[0703] Data processing: Generate notification messages and send them to the user terminal

[0704] Output: Notification sent to the user's device

[0705] Specific operation: The server uses a push notification service (e.g., Firebase Cloud Messaging) to send a notification to the user device.

[0706] Step 10:

[0707] The server performs big data analysis and predictions.

[0708] Input: Accumulated video data and abnormal behavior detection logs

[0709] Data processing: Analyze data and train predictive models using Hadoop or Spark

[0710] Output: Prediction results, optimal resource allocation proposals

[0711] Specific operation: The server processes the data using big data tools and trains a predictive model to predict future abnormal behavior.

[0712] (Application example 1)

[0713] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0714] In modern society, surveillance systems are important for ensuring public safety. However, current surveillance systems lack the ability to detect abnormal behavior and emergencies in real time, making it difficult for security guards and relevant agencies to respond immediately. Furthermore, they lack the functionality to quickly detect abnormalities on-site using mobile devices and respond appropriately, resulting in a decrease in the efficiency of security services.

[0715] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0716] In this invention, the server includes means for acquiring camera footage for identifying people or objects, means for analyzing camera footage data and identifying people or objects, means for detecting abnormal behavior or emergencies, means for automatically issuing an alarm based on the detected abnormal behavior or emergency, means for coordinating with relevant organizations in real time, means for displaying an alarm on a display device mounted on a mobile terminal, means for transmitting video data from the mobile terminal to the server, and means for transmitting analyzed data and alarm information to relevant organizations. This enables security guards to immediately recognize abnormal behavior on-site and respond quickly, strengthening cooperation with relevant organizations and ensuring public safety.

[0717] "Means for acquiring camera footage to identify people or objects" refers to the function of collecting video data using surveillance cameras.

[0718] "Means for analyzing camera video data and identifying people or objects" refers to algorithms or software for analyzing collected video data and identifying people or objects.

[0719] "Means for detecting abnormal behavior or emergency situations" refers to the ability to use AI or machine learning models to detect behavior that deviates from normal behavior or emergency situations in real time.

[0720] "Means for automatically issuing an alert based on detected abnormal behavior or emergency situations" refers to a function for automatically issuing an alert when abnormal behavior or emergency situations are detected.

[0721] "Means of coordinating with relevant agencies in real time" refers to a system for sharing information in real time with relevant agencies such as the police and fire department and taking joint action.

[0722] "Means for displaying an alert on a display device installed in a mobile terminal" refers to a function for displaying alerts or notification messages in real time on a display device such as a smartphone or smart glasses.

[0723] "Means for transmitting video data from a mobile terminal to a server" refers to a function for uploading video data from a mobile terminal to a server via Internet communication.

[0724] "Means for transmitting analyzed data and warning information to relevant organizations" refers to a system for instantly transmitting AI analysis results and warning notifications to designated relevant organizations.

[0725] MODE FOR CARRYING OUT THE INVENTION

[0726] The present invention provides a security service using a high-performance surveillance camera system and mobile devices that utilize AI. An embodiment of this system is described in detail below.

[0727] System Configuration

[0728] The system consists of the following main components:

[0729] 1. Surveillance camera (terminal)

[0730] 2. Mobile devices (e.g., smart glasses)

[0731] 3. Server

[0732] 4. User devices (smartphones and PCs)

[0733] 5. Related agencies (police, fire department, etc.)

[0734] Acquiring and transmitting camera images

[0735] Surveillance cameras: These cameras are installed in cities, commercial districts, and public facilities and have the ability to capture video in real time. These videos are compressed at regular intervals and sent to a server.

[0736] Receiving and analyzing video data

[0737] Server: Receives and decompresses video data sent from surveillance cameras. The data is then input into an AI model to detect abnormal behavior and emergencies. Examples of AI models used include YOLO and SSD.

[0738] Detection of abnormal behavior: When the server detects abnormal behavior or an emergency, it captures the timestamp and location information and records it in a database.

[0739] Alarm display on mobile device

[0740] Smart glasses: Alerts from the server and notifications about abnormal behavior are displayed on the mobile device in real time to warn the user.

[0741] Video data transmission: On-site video data is uploaded from the smart glasses or smartphone to a server for further analysis.

[0742] Cooperation with related organizations

[0743] Alert notification: The server generates an alert message based on the detected abnormal behavior and sends the analyzed data to the relevant authorities. The notification includes the video clip, timestamp, and location information.

[0744] Big Data Analysis and Prediction

[0745] Big data analysis: The server performs big data analysis using the accumulated video data and abnormal behavior logs, which makes it possible to predict and prevent crimes and emergencies.

[0746] Specific prompt examples

[0747] 1. Example 1: A person is in possession of a knife at a train station.

[0748] 2. Example 2: Theft is detected in a shopping mall.

[0749] Hardware and software used

[0750] Hardware:

[0751] Surveillance cameras: high-resolution cameras (e.g. CCTV cameras)

[0752] Mobile devices: smart glasses (e.g., Google Glass, Vuzix Blade), smartphones

[0753] software:

[0754] AI model: YOLO, SSD

[0755] Server: AWS, Google Cloud

[0756] Notification systems: Firebase, Twilio

[0757] Program processing

[0758] 1. The device (surveillance camera) captures video in real time and sends it to the server at regular intervals.

[0759] 2. The server receives the video data, analyzes it using an AI model, and detects abnormal behavior or emergencies.

[0760] 3. The server generates an alert based on the detected abnormal behavior or emergency situation and displays it on the mobile device (smart glasses, smartphone).

[0761] 4. Send the alert and detailed data to relevant authorities (police, fire department, etc.).

[0762] This will enable security guards to immediately recognize any abnormal behavior on-site and respond quickly, strengthening cooperation with relevant agencies and ensuring public safety.

[0763] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0764] Step 1:

[0765] The terminal (surveillance camera) captures video in real time and temporarily stores the video data in a buffer. At regular intervals, this video data is compressed and sent to the server. The input here is the raw video data captured by the camera, and the output is compressed video data.

[0766] Step 2:

[0767] The server receives the compressed video data sent from the terminal and decompresses it. The decompressed video data is ready for analysis frame by frame. The input here is the compressed video data and the output is the decompressed frame data.

[0768] Step 3:

[0769] The server inputs the decompressed video data into a generative AI model (e.g., YOLO or SSD) to identify people or objects. This identification data is then used to detect abnormal behavior or emergencies. Here, the input is the decompressed frame data, and the output is a list of identified objects.

[0770] Step 4:

[0771] The server analyzes the identified object list to detect anomalous behavior or emergencies. If an anomaly is detected, its timestamp and location information are recorded in a database. The input here is the list of identified objects, and the output is the detection of anomalous behavior or emergencies.

[0772] Step 5:

[0773] The server automatically issues an alert based on the detected abnormal behavior or emergency situation. This alert is sent to the display device of the mobile device (smart glasses or smartphone) and displays a warning to the user. The input here is the abnormal behavior detection result, and the output is the alert message.

[0774] Step 6:

[0775] A mobile device (smart glasses or smartphone) receives an alert message from the server and displays it to the user in real time. The user can then check the details of the alert and take necessary measures. The input here is the alert message, and the output is a notification display to the user.

[0776] Step 7:

[0777] The server sends the alert and detailed data (video clip, timestamp, location information) to the relevant authorities (police, fire department, etc.). The relevant authorities then take prompt action based on this information. The input here is the alert message and detailed data, and the output is a notification to the relevant authorities.

[0778] Step 8:

[0779] The server performs big data analysis and trains a predictive model based on the accumulated video data and abnormal behavior logs. This enables future predictions of crimes and emergencies and allows preventative measures to be planned. The input here is past video data and abnormal behavior logs, and the output is the results of the predictive model.

[0780] By taking the above steps, security guards can immediately recognize any abnormal behavior on-site and respond quickly, strengthening cooperation with relevant agencies and ensuring public safety.

[0781] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0782] The system of the present invention combines an emotion engine with a high-performance surveillance camera system that utilizes AI, and is capable of identifying people and objects, detecting abnormal behavior and emergency situations, automatically issuing alarms, and recognizing the user's emotions and responding based on those emotions. The following describes in detail the embodiments of the present invention.

[0783] System Configuration

[0784] The system includes the following main components:

[0785] 1. Terminal (surveillance camera)

[0786] 2. Server

[0787] 3. Emotion Engine

[0788] 4. User devices (smartphones and PCs)

[0789] 5. Related agencies (police, fire department, etc.)

[0790] Acquiring and transmitting camera images

[0791] The device (surveillance camera) captures real-time video of a specified area. The video is divided into frames and temporarily stored in a buffer. The video data in the buffer is compressed at regular intervals and sent to the server.

[0792] Receiving and analyzing video data

[0793] The server receives the video data sent from the device. The received data is decompressed and immediately input into the AI ​​model. This model identifies people and objects, and detects abnormal behavior and emergencies. The detected data is recorded in a database.

[0794] Detecting abnormal behavior and issuing alerts

[0795] The server captures and flags any abnormal behavior or emergency situations, capturing their timestamp and location information. The video clip, timestamp, and location information are saved and generated as an alert message, which is then sent to the appropriate authorities in the form of a push notification.

[0796] Introducing the Emotion Engine

[0797] The emotion engine uses the captured video data and abnormal behavior analysis data to recognize emotions from the user's facial expressions and voice. This emotion data is used to evaluate abnormal behavior and optimize responses.

[0798] Optimizing responses based on perceived emotions

[0799] The server uses the emotion data obtained from the emotion engine to optimize the content of the alert and emergency response. For example, if a user is in a panic state, the server sends an alert that increases the level of urgency to the relevant authorities.

[0800] User Notifications

[0801] User devices (smartphones and PCs) receive notifications when abnormal behavior or emergencies occur, including a detailed description of the event, a link to a video clip, and appropriate countermeasures based on the user's feelings.

[0802] Cooperation with related organizations

[0803] The server waits for confirmation of the alert sent to the relevant authorities, and once confirmation is received, it provides the next important update (e.g., changes in the situation on the ground, additional video clips, etc.) in real time, allowing the relevant authorities to respond quickly and accurately.

[0804] Big Data Analysis and Prediction

[0805] The server uses the accumulated video data and emotion data for big data analysis, training predictive models to forecast crimes and emergencies, and supporting optimal resource allocation and countermeasure planning based on the prediction results.

[0806] Specific examples

[0807] Example 1: Incident detection on public transport

[0808] The device (surveillance camera) continues to monitor the video near the ticket gate of the station. Among the people passing by, it detects one person holding a knife.

[0809] The server detects this abnormal behavior and uses an emotion engine to analyze the emotions of the person wielding the knife and those around them. If emotions of panic or fear are recognized, a high-level alert is generated and sent to the relevant authorities.

[0810] The relevant authorities (police) receive a notification and dispatch the nearest police officer to the scene. The server continues to monitor the footage from the camera and provides additional information to the relevant authorities.

[0811] The user (station staff) will receive a notification on their smartphone, and will be able to immediately grasp the situation and prepare a response.

[0812] Example 2: Suspicious behavior detection in a commercial area

[0813] The device (surveillance camera) monitors the video at the entrance of the shopping mall and detects the behavior of a person stealing goods from multiple stores.

[0814] The server detects this abnormal behavior and uses an emotion engine to analyze the emotions of the thief and those around them. If emotions of anxiety or confusion are recognized, an alert is generated suggesting appropriate responses and sent to the relevant authorities.

[0815] The relevant authorities (police) receive the notification and immediately dispatch the nearest police officers to the scene. The server updates the relevant authorities as the situation changes.

[0816] The user (store staff) will receive a notification on their smartphone, immediately check the security camera footage, and take appropriate action while waiting for the police to arrive.

[0817] In this way, by combining an emotion engine with an AI-powered surveillance camera system, abnormal behavior can be detected in real time, strengthening collaboration with users and relevant agencies, and enabling swift and accurate responses. By utilizing big data analysis and predictive models, crime prevention and efficient resource allocation become possible.

[0818] The processing flow will be explained below.

[0819] Step 1:

[0820] The device (surveillance camera) captures real-time video of the specified area. The video is divided into frames and temporarily stored in a buffer.

[0821] Step 2:

[0822] The terminal (surveillance camera) compresses the video data in the buffer at regular intervals (e.g., every second) and sends it to a server via the Internet.

[0823] Step 3:

[0824] The server receives the video data sent from the device, decompresses it, and immediately inputs it into the AI ​​model.

[0825] Step 4:

[0826] The server analyzes the transmitted video data and identifies people and objects using an AI model that is optimized based on pre-trained data.

[0827] Step 5:

[0828] Based on the analysis results, the server tracks the movements of people and objects in the video and uses an abnormal behavior detection algorithm to detect abnormal behavior such as falls, fights, or the possession of knives.

[0829] Step 6:

[0830] When the server detects abnormal or emergency activity, it clips the frame of data and flags it as abnormal activity. The video clip, timestamp, and location information are saved.

[0831] Step 7:

[0832] The server uses an emotion engine to analyze the user's facial expressions and voice based on the analysis results of the video data and abnormal behavior, and recognizes emotions.

[0833] Step 8:

[0834] The server then optimizes the content of the alert and emergency response based on the recognized emotion. For example, if a user is in a panic state, it will generate a high-urgency alert and send it to the relevant authorities.

[0835] Step 9:

[0836] The server sends an alert to the relevant authorities (police, fire department) in the form of a push notification. The alert message includes details of the incident, a video clip, a timestamp, and location information.

[0837] Step 10:

[0838] The relevant authorities (police and fire departments) receive an alert notification and begin preparing to respond to the incident, dispatching the nearest police officers and firefighters to the scene.

[0839] Step 11:

[0840] The server monitors the response status of relevant agencies and continues to transmit video data in real time if the situation at the scene changes, providing the latest information.

[0841] Step 12:

[0842] The server sends notifications of emergencies or abnormal behavior to specific user devices (e.g., commercial facility staff, security personnel), including detailed descriptions, links to video clips, and appropriate countermeasures based on the recognized emotions.

[0843] Step 13:

[0844] Users (such as commercial facility staff) receive notifications on their smartphones or PCs, immediately check the situation, and take action to ensure the safety of their surroundings and prevent damage, if necessary.

[0845] Step 14:

[0846] The server stores the accumulated video data and emotion data and uses it for big data analysis, using predictive models to forecast future crimes and emergencies.

[0847] Step 15:

[0848] The server generates detailed reports based on the analysis results to help optimize resource allocation and plan crime prevention activities, and the generated reports enable efficient countermeasures to be taken.

[0849] This enables the system to utilize AI for real-time monitoring and immediate response, and employs an emotion engine to take appropriate action in the event of abnormal behavior or an emergency, significantly improving safety in cities, commercial districts, public transportation, and communities.

[0850] Example 2

[0851] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0852] Current surveillance camera systems are limited to simply monitoring video footage and detecting abnormal behavior, and lack the ability to accurately assess users' emotions and situations and respond quickly and appropriately. Furthermore, their functionality for coordinating with relevant agencies and preventing abnormal behavior is limited. This leads to problems such as delayed initial responses to emergencies and difficulty allocating appropriate resources.

[0853] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0854] In this invention, the server includes a means for acquiring camera footage, a means for analyzing the acquired camera footage in real time and identifying people or objects, a means for detecting abnormal behavior or emergencies, a means for automatically issuing an alert based on the detected abnormal behavior or emergency, a means for coordinating with relevant organizations in real time, and a means for recognizing emotions from the user's facial expressions and voice and optimizing responses. This integrates the process from detecting abnormal behavior to emergency response, enabling flexible responses based on the user's emotions. Furthermore, the combined use of big data analysis and predictive models enables more accurate prediction and prevention of abnormal behavior and emergencies, enabling optimal resource allocation and rapid response.

[0855] "Camera footage" refers to visual information obtained from a surveillance camera.

[0856] "Video data" refers to data that represents camera images in digital format.

[0857] "Real-time analysis" refers to processing and analyzing data as soon as it is received.

[0858] "Person or object identification" refers to analyzing surveillance camera footage and identifying people or objects present in the footage.

[0859] "Abnormal behavior" refers to behavior that deviates from general patterns of behavior and is considered dangerous or suspicious.

[0860] An "emergency" is a sudden situation or event that requires immediate action.

[0861] "Alert" refers to a warning or notification issued when abnormal behavior or an emergency is detected.

[0862] "Related agencies" refers to external organizations (e.g., police, fire department, etc.) that respond to abnormal behavior or emergencies.

[0863] "Facial expressions" refer to the movements and changes of facial muscles that express a person's emotions and psychological state.

[0864] "Audio" refers to data related to human voices and sounds.

[0865] "Emotion recognition" refers to identifying a user's emotions and mental state from their facial expressions and voice.

[0866] "Optimizing responses" refers to selecting and implementing the most appropriate and effective response measures based on emotional data and the situation.

[0867] "Big data analytics" refers to the process of analyzing large amounts of data and extracting meaningful information and patterns.

[0868] "Predictive model" refers to a mathematical or statistical model used to forecast future events or trends based on historical data.

[0869] The system of the present invention combines an emotion engine with a high-performance surveillance camera system that utilizes AI, and is capable of identifying people and objects, detecting abnormal behavior and emergency situations, automatically issuing alarms, and recognizing the user's emotions and responding based on those emotions. The following are modes for implementing the present invention.

[0870] System Configuration

[0871] The system includes the following main components:

[0872] 1. Terminal (surveillance camera)

[0873] 2. Server

[0874] 3. Emotion Engine

[0875] 4. User devices (smartphones and PCs)

[0876] 5. Related agencies (police, fire department, etc.)

[0877] Acquiring and transmitting camera images

[0878] The device (surveillance camera) captures real-time video of a specified area. For example, this applies to surveillance cameras installed in surveillance systems for public transportation or commercial districts. The video is divided into frames and temporarily stored in a buffer. The video data is periodically compressed using a compression technology such as H.264 and sent to a server using a secure protocol such as HTTPS.

[0879] Receiving and analyzing video data

[0880] The server receives the compressed video data sent from the device. The received data is decompressed and input into a generative AI model (e.g., YOLOv4 or OpenPose). This AI model identifies people and objects and detects abnormal behavior and emergencies. For example, it can detect dangerous behavior such as someone carrying a knife on public transportation.

[0881] Detecting abnormal behavior and issuing alerts

[0882] The server then assigns a timestamp and location information to any abnormal behavior or emergency detected by the AI ​​model. Based on this, an alert message is generated and sent to the relevant authorities in the form of a push notification, along with a detailed description of the incident and a link to a video clip. For example, police and fire departments are notified immediately.

[0883] Introducing an emotion engine and optimizing responses

[0884] The emotion engine analyzes the user's facial expressions and voice from the acquired video data and abnormal behavior analysis data to recognize their emotions. The server can use this emotion data to optimize the content of warnings and emergency responses. For example, if a user is in a panic state, it can send a high-level alert to the relevant authorities.

[0885] User notification and collaboration with relevant organizations

[0886] User devices (smartphones and PCs) receive notifications when abnormal behavior or emergencies occur. The notifications include a detailed description of the event, a link to a video clip, and appropriate countermeasures based on emotion data, allowing users to respond quickly.

[0887] Big Data Analysis and Prediction

[0888] The server then performs big data analysis on the accumulated video and emotion data, which can then be used to train predictive models to forecast future abnormal behavior and emergencies. The prediction results can then be used to optimize resource allocation and support countermeasure planning.

[0889] Resulting System Behavior

[0890] This system analyzes real-time footage from surveillance cameras, detects abnormal behavior and emergencies, and immediately coordinates with relevant agencies to take appropriate action. Furthermore, it has the ability to analyze user emotions and respond appropriately based on those emotions, enabling rapid and effective response to emergencies. Big data analysis also improves prediction accuracy, enabling advance countermeasures and optimal resource allocation.

[0891] Examples of prompt statements

[0892] "Please explain in detail the process of the system that detects abnormal behavior at station ticket gates and also analyzes the emotions of those around you."

[0893] "Please explain a system that detects suspicious behavior in a shopping mall and performs sentiment analysis based on that behavior."

[0894] In this way, the system of the present invention enables real-time detection of abnormal behavior, prompt notification to users and relevant organizations, and optimal response based on emotional data.

[0895] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0896] Step 1:

[0897] The terminal (surveillance camera) acquires real-time video of the designated area.

[0898] Input: Video of the specified area

[0899] Processing: Capture video in frames (e.g. 30 frames per second) and save as frame data

[0900] Output: Frame data

[0901] Step 2:

[0902] The terminal (surveillance camera) stores the acquired frame data in a buffer, compresses it at regular intervals, and sends it to the server.

[0903] Input: Frame data

[0904] Processing: Frame data is temporarily stored in a buffer and compressed using a compression algorithm such as H.264

[0905] Output: Compressed video data is sent to the server

[0906] Step 3:

[0907] The server receives the compressed video data sent from the terminal and decompresses it.

[0908] Input: Compressed video data

[0909] Processing: Decompresses received data and restores it as frame data

[0910] Output: Decompressed frame data

[0911] Step 4:

[0912] The server inputs the decompressed frame data into a generative AI model to identify people or objects.

[0913] Input: Decompressed frame data

[0914] Processing: Input frame data into a generative AI model (e.g., YOLOv4, OpenPose, etc.) to identify people and objects.

[0915] Output: Data of identified people or objects (location, label, etc.)

[0916] Step 5:

[0917] The server analyzes the data of the identified person or object to detect abnormal behavior or emergency situations.

[0918] Input: Identified person or object data

[0919] Processing: Uses analytical algorithms to detect abnormal behavior and emergencies

[0920] Output: Data on detected abnormal behavior or emergency (e.g., timestamp, location, behavior type, etc.)

[0921] Step 6:

[0922] The server issues alerts based on detected abnormal behavior or emergency situations and sends push notifications to relevant authorities.

[0923] Input: Data on abnormal behavior and emergencies

[0924] Processing: Generate an alert message and send it to the relevant authorities in the form of a push notification.

[0925] Output: Alert message to relevant organizations

[0926] Step 7:

[0927] The emotion engine uses the acquired video data and abnormal behavior analysis data to recognize emotions from the user's facial expressions and voice.

[0928] Input: Video data, abnormal behavior analysis data

[0929] Processing: Apply emotion recognition algorithms to identify the user's emotions

[0930] Output: Recognized emotion data (e.g., panic, fear, anxiety, etc.)

[0931] Step 8:

[0932] The server uses the recognized emotion data to optimize alert content and emergency response.

[0933] Input: Emotion data

[0934] Processing: Adjust the content and urgency of the warning message and optimize the necessary response measures.

[0935] Output: Optimized warning messages and countermeasures

[0936] Step 9:

[0937] User devices (smartphones and PCs) receive notifications when abnormal behavior or emergencies occur.

[0938] Input: Optimized warning message

[0939] Action: Display a notification and inform the user of the situation

[0940] Output: A message to inform the user

[0941] Step 10:

[0942] The server waits for confirmation of receipt of alerts sent to relevant authorities and provides important updates in real time.

[0943] Input: Acknowledgments from relevant agencies and new video data

[0944] Processing: Get acknowledgements and real-time updates on important information like changing conditions on the ground and additional footage clips

[0945] Output: Update information for relevant organizations

[0946] Step 11:

[0947] The server uses the accumulated video data and emotion data for big data analysis.

[0948] Input: Accumulated video data, emotion data

[0949] Processing: Data is stored in a data warehouse and analyzed using data mining techniques

[0950] Output: Analysis results and training data for predictive models

[0951] Step 12:

[0952] The server uses predictive models to forecast crimes and emergencies and help plan countermeasures.

[0953] Input: Prediction model, analysis results

[0954] Processing: Applying predictive algorithms to predict future abnormal behavior and emergencies

[0955] Output: Prediction results and supporting information for action plans

[0956] (Application example 2)

[0957] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0958] Conventional surveillance systems have the ability to identify people and objects and detect abnormal behavior and emergencies, but they are unable to recognize user emotions and provide responses based on those emotions. Furthermore, responses after detecting abnormal behavior are limited, often resulting in failure to contact relevant authorities or propose appropriate countermeasures. To address these issues, the present invention aims to provide a system that adds emotion recognition functionality and optimizes responses after detecting abnormal behavior.

[0959] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for acquiring camera images for identifying people or objects, means for analyzing camera image data and identifying people or objects, means for detecting abnormal behavior or emergency situations, and means for recognizing user emotions and optimizing responses based on the emotions. This makes it possible to take appropriate action based not only on identification but also on emotion data when abnormal behavior or an emergency occurs.

[0960] "Person or object identification" refers to technology that analyzes camera video data to identify people or objects.

[0961] "Unusual behavior or emergency" means any behavior that deviates from normal patterns of behavior or any situation that poses an immediate danger to life or property.

[0962] "Automatic alerting means" means a technology that automatically issues an alert in accordance with pre-defined conditions based on detected abnormal behavior or emergency situations.

[0963] "Means for recognizing the user's emotions and optimizing the response based on that" refers to technology that analyzes the user's emotions using acquired video and audio data and provides the optimal response based on the results.

[0964] "Means of coordinating with relevant agencies in real time" refers to technology that allows for instant communication with relevant agencies such as the police and fire department, providing necessary information, and maintaining a cooperative relationship.

[0965] "Big data analytics and predictive models" are techniques for analyzing large amounts of data to derive specific patterns and trends, which can be used to predict abnormal behavior and emergencies.

[0966] "Means for providing video clips, timestamps, and location information" refers to technology that, when issuing an alert, adds detailed information such as specific video data, the time of occurrence, and location.

[0967] "Means for proposing appropriate countermeasures" refers to technology that presents optimal and effective countermeasures based on detected abnormal behavior and user emotional data.

[0968] System Overview

[0969] The present invention is a system that identifies people and objects, detects abnormal behavior and emergencies, and automatically issues an alarm. Furthermore, it recognizes the user's emotions and provides an optimal response based on those emotions, making the response after abnormal behavior is detected more effective.

[0970] System Components

[0971] The system includes the following major components:

[0972] 1. Surveillance cameras

[0973] 2. Server

[0974] 3. Emotion Recognition Engine

[0975] 4. User devices (smartphones and PCs)

[0976] 5. Related agencies (police, fire department, etc.)

[0977] System Operation

[0978] Acquiring and transmitting camera images

[0979] The surveillance camera captures real-time video of a specified area. The video is divided into frames and temporarily stored in a buffer. The video data in the buffer is compressed at regular intervals and sent to the server.

[0980] Receiving and analyzing video data

[0981] The server receives video data transmitted from the surveillance cameras. The received data is decompressed and immediately input into a generative AI model. This model identifies people and objects, and detects abnormal behavior and emergencies. The detected data is then recorded in a database.

[0982] Detecting abnormal behavior and issuing alerts

[0983] The server captures and flags any abnormal behavior or emergency situations, capturing their timestamp and location information. The video clip, timestamp, and location information are saved, and an alert message is generated and sent to the appropriate authorities in the form of a push notification.

[0984] Using the Emotion Recognition Engine

[0985] The server uses the acquired video data and abnormal behavior analysis data to recognize emotions from the user's facial expressions and voice using an emotion recognition engine. This emotion data is used to evaluate abnormal behavior and optimize responses.

[0986] Optimizing responses based on perceived emotions

[0987] The server uses the emotion data obtained from the emotion recognition engine to optimize the content of the alert and emergency response, for example, sending an alert to the relevant authorities that increases the level of urgency if the user is in a panic.

[0988] User notification and response

[0989] The user device receives a notification when abnormal behavior or an emergency occurs. The notification includes a detailed description of the event, a link to a video clip, and appropriate countermeasures based on the user's emotions. The notification arrives on the user's smartphone, allowing the user to immediately understand the situation and prepare a response.

[0990] Hardware and software used

[0991] Hardware:

[0992] Surveillance cameras (IP cameras, etc.)

[0993] Smartphone or PC

[0994] software:

[0995] OpenCV: Camera image acquisition and analysis

[0996] Keras: Loading and predicting emotion recognition models

[0997] requests: Server alert notification

[0998] Specific processing examples

[0999] Example 1: Suspicious behavior detection in a shopping mall

[1000] 1. A surveillance camera monitors footage at the entrance to a shopping mall and detects an individual stealing items from multiple stores.

[1001] 2. The server detects this abnormal behavior and uses an emotion recognition engine to analyze the emotions of the thief and those around them. If emotions of anxiety or confusion are recognized, an alert is generated suggesting appropriate responses and sent to the relevant authorities.

[1002] 3. The relevant authorities (police) receive the notification and immediately dispatch the nearest police officers to the scene. The server updates the relevant authorities with information as the situation changes.

[1003] 4. The user (store staff) receives a notification on their smartphone, immediately checks the security camera footage, and takes appropriate action while waiting for the police to arrive.

[1004] Prompt Sentence Examples

[1005] "Detect anomalous behavior in this camera footage and recognize the emotions of people around it. If anomalous behavior is detected, generate an alert message suggesting appropriate countermeasures based on the emotional data."

[1006] The above is an embodiment of the present invention. The present invention makes it possible to detect abnormal behavior in real time, strengthen cooperation with users and related organizations, and realize a prompt and accurate response.

[1007] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1008] Step 1:

[1009] The surveillance camera captures real-time video of a specified area. The input is the video data from the surveillance camera, and the output is the video frame stored in the buffer. This frame is temporarily stored for further processing.

[1010] Step 2:

[1011] The server compresses the video data in the buffer at regular intervals and sends it to the server. The input is the video frames stored in the buffer, and the output is the compressed video data. Data compression enables efficient use of network bandwidth.

[1012] Step 3:

[1013] The server receives the compressed video data and decompresses it. The input is the compressed video data, and the output is the decompressed video data. The decompressed data is immediately input into the generative AI model.

[1014] Step 4:

[1015] The server uses a generative AI model to identify people and objects. The input is the decompressed video data, and the output is the information of the identified people and objects. The generative AI model identifies people and objects based on video analysis.

[1016] Step 5:

[1017] The server further analyzes the information of the identified people and objects to detect abnormal behavior or emergency situations. The input is the information of the identified people and objects, and the output is a flag indicating abnormal behavior or emergency situation. An algorithm is used to detect patterns of abnormal behavior or emergency situations.

[1018] Step 6:

[1019] When an abnormal behavior or emergency is detected, the server acquires the timestamp and location information and sets a flag. The input is the flag for abnormal behavior detection or emergency detection, and the output is the timestamp and location information.

[1020] Step 7:

[1021] The server generates an alert message based on the video clip, timestamp, and location information, and sends it to the relevant organizations. The input is the timestamp, location information, and video clip, and the output is the generated alert message. It is sent to the relevant organizations in the form of a push notification.

[1022] Step 8:

[1023] The server uses an emotion recognition engine to analyze the user's facial expressions and voice based on the acquired video data and recognizes their emotions. The input is video data and the output is emotion data. The emotion recognition engine estimates the user's emotions from their facial expressions and voice.

[1024] Step 9:

[1025] The server optimizes the content of the alert and emergency response using the emotion data obtained from the emotion recognition engine. The input is emotion data, and the output is optimized alert content and emergency response measures. The urgency and content of the response are adjusted based on the emotion data.

[1026] Step 10:

[1027] The user device receives notifications when abnormal behavior or an emergency occurs. The input is the optimized alert content and emergency response measures, and the output is a notification displayed on the user device. The user receives the notification on their smartphone, allowing them to immediately understand the situation and prepare a response.

[1028] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1029] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1030] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[1031] [Third embodiment]

[1032] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1033] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[1034] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1035] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[1036] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1037] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1038] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1039] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1040] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1041] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1042] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1043] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[1044] The system of the present invention is a high-performance surveillance camera system that utilizes AI, and can identify people and objects, detect abnormal behavior and emergency situations, and automatically issue an alarm. The following describes in detail an embodiment of the present invention.

[1045] System Configuration

[1046] The system includes the following main components:

[1047] 1. Terminal (surveillance camera)

[1048] 2. Server

[1049] 3. User device (smartphone or PC)

[1050] 4. Related agencies (police, fire department, etc.)

[1051] Acquiring and transmitting camera images

[1052] The devices (surveillance cameras) are placed in cities, commercial districts, public transportation facilities, and communities. These cameras capture images in real time and store them in a buffer frame by frame. The video data in the buffer is compressed and sent to a server at regular intervals.

[1053] Receiving and analyzing video data

[1054] The server receives the video data sent from the device. The received data is decompressed and input into an AI model. This model identifies people and objects, and detects abnormal behavior and emergencies. For example, it can detect situations where a person falls or a fight breaks out. The detected data is recorded in a database.

[1055] Detecting abnormal behavior and issuing alerts

[1056] When an abnormal behavior or emergency situation is detected, the server acquires the timestamp and location information and prepares it for the emergency notification system. The alert is generated as an alert message containing the video clip, timestamp, and location information, and is sent to the relevant authorities in the form of a push notification.

[1057] Cooperation with related organizations

[1058] The server waits for confirmation of the alert sent to the relevant authorities, and once confirmation is received, it provides the next important update (e.g., changes in the situation on the ground, additional video clips, etc.) in real time, allowing the relevant authorities to respond quickly and accurately.

[1059] User Notifications

[1060] The server sends notifications to specific user devices when an emergency or abnormal behavior occurs, including a detailed description of the event and a link to a video clip, allowing users to take immediate action.

[1061] Big Data Analysis and Prediction

[1062] The server performs big data analysis using accumulated video data and abnormal behavior detection logs. It trains predictive models to forecast crimes and emergencies. Based on the prediction results, it helps optimize resource allocation and plan countermeasures. This improves the efficiency of crime prevention and emergency response.

[1063] Specific examples

[1064] Example 1: Incident detection on public transport

[1065] The device (surveillance camera) continues to monitor the video near the ticket gate of the station. Among the people passing by, it detects one person holding a knife.

[1066] The server detects this anomalous behavior and sends a notification to the emergency notification system, which includes the video clip, a timestamp, and location information.

[1067] The relevant authorities (police) receive a notification and dispatch the nearest police officer to the scene. The server continues to monitor the footage from the camera and provides additional information to the relevant authorities.

[1068] The user (station staff) will receive a notification on their smartphone, and will be able to immediately grasp the situation and prepare a response.

[1069] Example 2: Suspicious behavior detection in a commercial area

[1070] The device (surveillance camera) monitors the video at the entrance of the shopping mall and detects the behavior of a person stealing goods from multiple stores.

[1071] The server detects this unusual behavior and sends an alert to the police, along with video clips and location information, for rapid response.

[1072] The relevant authorities (police) receive the notification and immediately dispatch the nearest police officers to the scene. The server updates the relevant authorities as the situation changes.

[1073] The user (store staff) will receive a notification on their smartphone, immediately check the security camera footage, and take appropriate action while waiting for the police to arrive.

[1074] This allows the AI-powered surveillance camera system to detect abnormal behavior in real time, strengthen cooperation with relevant agencies, and enable rapid response. It also utilizes big data analytics and predictive models to prevent crime and efficiently allocate resources.

[1075] The processing flow will be explained below.

[1076] Step 1:

[1077] The device (surveillance camera) captures real-time video of the specified area. The video is divided into frames and temporarily stored in a buffer.

[1078] Step 2:

[1079] The terminal (surveillance camera) compresses the video data in the buffer at regular intervals (e.g., every second) and sends it to a server via the Internet.

[1080] Step 3:

[1081] The server receives the video data sent from the device, decompresses it, and immediately inputs it into the AI ​​model.

[1082] Step 4:

[1083] The server analyzes the transmitted video data and identifies people and objects using an AI model that is optimized based on pre-trained data.

[1084] Step 5:

[1085] Based on the analysis results, the server tracks the movements of people and objects in the video and uses an abnormal behavior detection algorithm to detect abnormal behavior such as falls, fights, or the possession of knives.

[1086] Step 6:

[1087] When the server detects abnormal or emergency activity, it clips the frame of data and flags it as abnormal activity. The video clip, timestamp, and location information are saved.

[1088] Step 7:

[1089] If any unusual behavior or emergency is flagged, the server immediately generates an alert message, which includes details of the incident, video clips, timestamps, and location information.

[1090] Step 8:

[1091] The server sends an alert to the relevant authorities (police and fire departments) in the form of a push notification. After sending the alert, it waits for confirmation of receipt of the alert, and once confirmation is received, it proceeds to the next step.

[1092] Step 9:

[1093] The relevant authorities (police and fire departments) receive an alert notification and begin preparing to respond at the scene, dispatching the nearest police officers and firefighters to the scene.

[1094] Step 10:

[1095] The server monitors the response status of relevant agencies and continues to transmit video data in real time if the situation at the scene changes, providing the latest information.

[1096] Step 11:

[1097] The server sends notifications of emergencies or unusual behavior to specific user devices (e.g., commercial facility staff, security personnel), including detailed descriptions and links to video clips.

[1098] Step 12:

[1099] Users (such as commercial facility staff) receive notifications on their smartphones or PCs, immediately check the situation, and take action to ensure the safety of their surroundings and prevent damage, if necessary.

[1100] Step 13:

[1101] The server stores the accumulated video data and abnormal behavior detection logs, and uses them for big data analysis, using predictive models to forecast future crimes and emergencies.

[1102] Step 14:

[1103] The server generates detailed reports based on the analysis results to help optimize resource allocation and plan crime prevention activities, and the generated reports enable efficient countermeasures to be taken.

[1104] This enables the system to leverage AI for real-time monitoring and immediate response, significantly improving safety in cities, commercial districts, public transportation, and communities.

[1105] Example 1

[1106] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1107] Conventional surveillance camera systems have difficulty detecting abnormal behavior and emergencies in real time and providing prompt notification. Furthermore, due to insufficient collaboration with relevant agencies and insufficient predictive capabilities based on big data analysis, there were issues with the efficiency of crime prevention and emergency response. Furthermore, the notification functionality was limited, preventing users from taking immediate action.

[1108] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1109] In this invention, the server includes a device for acquiring camera footage, a device for compressing the acquired video data and sending it to the server, a device for decompressing the data received by the server and inputting it into an AI model to identify people or objects, a device for detecting abnormal behavior or emergencies, a device for recording information on the detected abnormal behavior or emergency in a database and generating an alert, a device for coordinating with relevant organizations in real time, and a device for sending notifications of emergency situations or abnormal behavior to specific user devices. This enables real-time detection of abnormal behavior or emergencies and rapid notification, strengthens cooperation with relevant organizations, and enables crime prevention and efficient resource allocation through predictions using big data analysis.

[1110] A "device for acquiring camera images" is a device that has the function of capturing images in real time in the environment in which it is installed and storing the image data.

[1111] The "device that compresses video data and transmits it to a server" is a device that has the function of compressing acquired video data and transmitting it to a server in an efficient format.

[1112] "Device that decompresses received data and inputs it into an AI model" refers to a device that has the function of decompressing compressed data received by the server and inputting it into an AI model.

[1113] A "device for identifying people or objects" is a device that has the function of identifying people or objects in video data using an AI model.

[1114] The "device for detecting abnormal behavior or emergency situations" is a device that has the function of detecting abnormal behavior or emergency situations that exceed a predetermined standard based on identified data.

[1115] An "alert generating device" is a device that has the function of generating and notifying an alert based on information on detected abnormal behavior or emergency situations.

[1116] The "device that cooperates with relevant organizations in real time" is a device that has the function of notifying relevant organizations of generated alerts in real time and ensuring cooperation.

[1117] A "device that transmits a notification to a specific user device" is a device that has a function of transmitting a notification to a specific user device when abnormal behavior or an emergency occurs.

[1118] A "device that uses big data analysis and predictive models" is a device that has the function of predicting future abnormal behavior or emergency situations using big data analysis and predictive models based on accumulated video data and abnormal behavior detection logs.

[1119] A "database recording device" is a device that has the function of storing information on detected abnormal behavior or emergency situations in a database.

[1120] MODE FOR CARRYING OUT THE INVENTION

[1121] The present invention relates to a high-performance surveillance camera system that utilizes AI. Hereinafter, an embodiment of the present invention will be described in detail.

[1122] System Configuration

[1123] The system mainly consists of the following components:

[1124] 1. Terminal (surveillance camera)

[1125] 2. Server

[1126] 3. User device (smartphone or PC)

[1127] 4. Related agencies (police, fire department, etc.)

[1128] Acquiring camera footage

[1129] The devices (surveillance cameras) are placed in cities, commercial districts, public transportation facilities, and communities. These cameras capture video in real time and store it in a buffer frame by frame. This stored video data is recompressed at regular intervals and sent to a server.

[1130] Video data compression and transmission

[1131] The device compresses the captured video data using a compression method such as H.264. The compressed data is then encrypted and securely sent to the server using the HTTPS protocol.

[1132] Receiving and decompressing video data

[1133] The server receives the compressed data sent from the device, stores it temporarily, and then decompresses it using decompression software such as FFmpeg to restore the original video.

[1134] Identification and detection with AI models

[1135] The server uses a deep learning framework (e.g., TensorFlow or PyTorch) to input the decompressed video data into an AI model, which then identifies people and objects and detects abnormal behavior and emergencies.

[1136] Recording abnormal behavior and generating alerts

[1137] The server records the detected abnormal behavior and emergency information in a database (e.g., MySQL, PostgreSQL), and simultaneously generates an alert message and sends it to the emergency notification system.

[1138] Sending and acknowledging alerts

[1139] The server sends the generated alert message to the relevant organizations in the form of a push notification. After sending, it waits for an ACK (acknowledgment) and, once it receives the ACK, provides additional information in real time.

[1140] User Notifications

[1141] The server notifies specific user devices of emergencies or abnormal behavior, including a detailed description of the event and a link to a video clip, allowing users to take immediate action.

[1142] Big Data Analysis and Prediction

[1143] The server uses the accumulated video data and abnormal behavior detection logs to perform big data analysis and predictive models. It processes the data using big data processing tools such as Hadoop and Spark to train predictive models. This makes it possible to predict crimes and emergencies, and helps with optimal resource allocation and countermeasure planning.

[1144] Specific examples

[1145] Example 1: Incident detection on public transport

[1146] The device (surveillance camera) continues to monitor the video near the ticket gate of the station. Among the people passing by, it detects one person holding a knife.

[1147] The server detects this unusual behavior and sends a notification to the emergency notification system, which includes a video clip, a timestamp, and location information.

[1148] The relevant authorities (police) are notified and dispatch the nearest police officers to the scene. The server continues to monitor the footage from the camera and provides additional information to the relevant authorities.

[1149] The user (station staff) will receive a notification on their smartphone, and will be able to immediately grasp the situation and prepare a response.

[1150] Prompt Sentence Examples

[1151] "Please describe an AI model that instantly detects a person with a knife at a station ticket gate and sends a notification to the appropriate authorities via an emergency notification system."

[1152] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1153] Step 1:

[1154] The terminals (surveillance cameras) capture images in real time in the areas where they are installed.

[1155] Input: Video data from a surveillance camera sensor

[1156] Data processing: Capture video frame by frame

[1157] Output: Video data divided into frames is stored in a buffer

[1158] Specific operation: The surveillance camera captures 30 frames of video per second and stores them in buffer memory.

[1159] Step 2:

[1160] The terminal compresses the video data in the buffer at regular intervals.

[1161] Input: Buffered raw video data

[1162] Data processing: Compresses video using compression algorithms such as H.264

[1163] Output: Compressed video data

[1164] Specific operation: Every 10 seconds, the device compresses the video data stored in the buffer and stores it in the output buffer.

[1165] Step 3:

[1166] The terminal transmits the compressed video data to the server.

[1167] Input: Compressed video data

[1168] Data processing: Encrypting data and preparing it for transmission using the HTTPS protocol

[1169] Output: Secure compressed video data sent to the server

[1170] Specific operation: The terminal encrypts the compressed video data using HTTPS and sends it to the server.

[1171] Step 4:

[1172] The server receives the compressed data sent from the terminal.

[1173] Input: Compressed and encrypted video data

[1174] Data processing: Decrypt the received data and store it in temporary storage

[1175] Output: Compressed data saved to storage

[1176] Specific operation: The server receives data sent via HTTPS, decrypts it, and stores it in storage.

[1177] Step 5:

[1178] The server decompresses the received data.

[1179] Input: Compressed video data

[1180] Data processing: Decompress the compressed data using decompression software such as FFmpeg.

[1181] Output: Original video data

[1182] Specific operation: The server uses FFmpeg to decompress the compressed data and return it to the original video frames.

[1183] Step 6:

[1184] The server inputs the decompressed video data into the AI ​​model.

[1185] Input: Decompressed video data

[1186] Data processing: Analyze video using deep learning frameworks such as TensorFlow and PyTorch

[1187] Output: Recognition results (identification of people and objects, detection of abnormal behavior)

[1188] How it works: The AI ​​model inputs video data frame by frame to detect abnormal behavior or emergencies.

[1189] Step 7:

[1190] The server records any abnormal behavior or emergency situations and generates alerts.

[1191] Input: Recognition results by AI model

[1192] Data processing: Recording abnormal behavior and emergency information in a database and generating alerts

[1193] Output: Data recorded in the database, generated alert messages

[1194] Specific operation: The server records abnormal behavior information in MySQL or PostgreSQL and generates an alert.

[1195] Step 8:

[1196] The server sends an alert message to the appropriate authority and waits for an acknowledgement.

[1197] Input: The generated alert message

[1198] Data processing: Generate and send a push notification containing the alert message

[1199] Output: Notification sent to the relevant authorities, acknowledgement (ACK)

[1200] Specific operation: The server sends a notification using REST API or WebSocket and waits for confirmation of receipt.

[1201] Step 9:

[1202] The server sends a notification to the specific user terminal.

[1203] Input: Information about abnormal behavior or emergency situations

[1204] Data processing: Generate notification messages and send them to the user terminal

[1205] Output: Notification sent to the user's device

[1206] Specific operation: The server uses a push notification service (e.g., Firebase Cloud Messaging) to send a notification to the user device.

[1207] Step 10:

[1208] The server performs big data analysis and predictions.

[1209] Input: Accumulated video data and abnormal behavior detection logs

[1210] Data processing: Analyze data and train predictive models using Hadoop or Spark

[1211] Output: Prediction results, optimal resource allocation proposals

[1212] Specific operation: The server processes the data using big data tools and trains a predictive model to predict future abnormal behavior.

[1213] (Application example 1)

[1214] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1215] In modern society, surveillance systems are important for ensuring public safety. However, current surveillance systems lack the ability to detect abnormal behavior and emergencies in real time, making it difficult for security guards and relevant agencies to respond immediately. Furthermore, they lack the functionality to quickly detect abnormalities on-site using mobile devices and respond appropriately, resulting in a decrease in the efficiency of security services.

[1216] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1217] In this invention, the server includes means for acquiring camera footage for identifying people or objects, means for analyzing camera footage data and identifying people or objects, means for detecting abnormal behavior or emergencies, means for automatically issuing an alarm based on the detected abnormal behavior or emergency, means for coordinating with relevant organizations in real time, means for displaying an alarm on a display device mounted on a mobile terminal, means for transmitting video data from the mobile terminal to the server, and means for transmitting analyzed data and alarm information to relevant organizations. This enables security guards to immediately recognize abnormal behavior on-site and respond quickly, strengthening cooperation with relevant organizations and ensuring public safety.

[1218] "Means for acquiring camera footage to identify people or objects" refers to the function of collecting video data using surveillance cameras.

[1219] "Means for analyzing camera video data and identifying people or objects" refers to algorithms or software for analyzing collected video data and identifying people or objects.

[1220] "Means for detecting abnormal behavior or emergency situations" refers to the ability to use AI or machine learning models to detect behavior that deviates from normal behavior or emergency situations in real time.

[1221] "Means for automatically issuing an alert based on detected abnormal behavior or emergency situations" refers to a function for automatically issuing an alert when abnormal behavior or emergency situations are detected.

[1222] "Means of coordinating with relevant agencies in real time" refers to a system for sharing information in real time with relevant agencies such as the police and fire department and taking joint action.

[1223] "Means for displaying an alert on a display device installed in a mobile terminal" refers to a function for displaying alerts or notification messages in real time on a display device such as a smartphone or smart glasses.

[1224] "Means for transmitting video data from a mobile terminal to a server" refers to a function for uploading video data from a mobile terminal to a server via Internet communication.

[1225] "Means for transmitting analyzed data and warning information to relevant organizations" refers to a system for instantly transmitting AI analysis results and warning notifications to designated relevant organizations.

[1226] MODE FOR CARRYING OUT THE INVENTION

[1227] The present invention provides a security service using a high-performance surveillance camera system and mobile devices that utilize AI. An embodiment of this system is described in detail below.

[1228] System Configuration

[1229] The system consists of the following main components:

[1230] 1. Surveillance camera (terminal)

[1231] 2. Mobile devices (e.g., smart glasses)

[1232] 3. Server

[1233] 4. User devices (smartphones and PCs)

[1234] 5. Related agencies (police, fire department, etc.)

[1235] Acquiring and transmitting camera images

[1236] Surveillance cameras: These cameras are installed in cities, commercial districts, and public facilities and have the ability to capture video in real time. These videos are compressed at regular intervals and sent to a server.

[1237] Receiving and analyzing video data

[1238] Server: Receives and decompresses video data sent from surveillance cameras. The data is then input into an AI model to detect abnormal behavior and emergencies. Examples of AI models used include YOLO and SSD.

[1239] Detection of abnormal behavior: When the server detects abnormal behavior or an emergency, it captures the timestamp and location information and records it in a database.

[1240] Alarm display on mobile device

[1241] Smart glasses: Alerts from the server and notifications about abnormal behavior are displayed on the mobile device in real time to warn the user.

[1242] Video data transmission: On-site video data is uploaded from the smart glasses or smartphone to a server for further analysis.

[1243] Cooperation with related organizations

[1244] Alert notification: The server generates an alert message based on the detected abnormal behavior and sends the analyzed data to the relevant authorities. The notification includes the video clip, timestamp, and location information.

[1245] Big Data Analysis and Prediction

[1246] Big data analysis: The server performs big data analysis using the accumulated video data and abnormal behavior logs, which makes it possible to predict and prevent crimes and emergencies.

[1247] Specific prompt examples

[1248] 1. Example 1: A person is in possession of a knife at a train station.

[1249] 2. Example 2: Theft is detected in a shopping mall.

[1250] Hardware and software used

[1251] Hardware:

[1252] Surveillance cameras: high-resolution cameras (e.g. CCTV cameras)

[1253] Mobile devices: smart glasses (e.g., Google Glass, Vuzix Blade), smartphones

[1254] software:

[1255] AI model: YOLO, SSD

[1256] Server: AWS, Google Cloud

[1257] Notification systems: Firebase, Twilio

[1258] Program processing

[1259] 1. The device (surveillance camera) captures video in real time and sends it to the server at regular intervals.

[1260] 2. The server receives the video data, analyzes it using an AI model, and detects abnormal behavior or emergencies.

[1261] 3. The server generates an alert based on the detected abnormal behavior or emergency situation and displays it on the mobile device (smart glasses, smartphone).

[1262] 4. Send the alert and detailed data to relevant authorities (police, fire department, etc.).

[1263] This will enable security guards to immediately recognize any abnormal behavior on-site and respond quickly, strengthening cooperation with relevant agencies and ensuring public safety.

[1264] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1265] Step 1:

[1266] The terminal (surveillance camera) captures video in real time and temporarily stores the video data in a buffer. At regular intervals, this video data is compressed and sent to the server. The input here is the raw video data captured by the camera, and the output is compressed video data.

[1267] Step 2:

[1268] The server receives the compressed video data sent from the terminal and decompresses it. The decompressed video data is ready for analysis frame by frame. The input here is the compressed video data and the output is the decompressed frame data.

[1269] Step 3:

[1270] The server inputs the decompressed video data into a generative AI model (e.g., YOLO or SSD) to identify people or objects. This identification data is then used to detect abnormal behavior or emergencies. Here, the input is the decompressed frame data, and the output is a list of identified objects.

[1271] Step 4:

[1272] The server analyzes the identified object list to detect anomalous behavior or emergencies. If an anomaly is detected, its timestamp and location information are recorded in a database. The input here is the list of identified objects, and the output is the detection of anomalous behavior or emergencies.

[1273] Step 5:

[1274] The server automatically issues an alert based on the detected abnormal behavior or emergency situation. This alert is sent to the display device of the mobile device (smart glasses or smartphone) and displays a warning to the user. The input here is the abnormal behavior detection result, and the output is the alert message.

[1275] Step 6:

[1276] A mobile device (smart glasses or smartphone) receives an alert message from the server and displays it to the user in real time. The user can then check the details of the alert and take necessary measures. The input here is the alert message, and the output is a notification display to the user.

[1277] Step 7:

[1278] The server sends the alert and detailed data (video clip, timestamp, location information) to the relevant authorities (police, fire department, etc.). The relevant authorities then take prompt action based on this information. The input here is the alert message and detailed data, and the output is a notification to the relevant authorities.

[1279] Step 8:

[1280] The server performs big data analysis and trains a predictive model based on the accumulated video data and abnormal behavior logs. This enables future predictions of crimes and emergencies and allows preventative measures to be planned. The input here is past video data and abnormal behavior logs, and the output is the results of the predictive model.

[1281] By taking the above steps, security guards can immediately recognize any abnormal behavior on-site and respond quickly, strengthening cooperation with relevant agencies and ensuring public safety.

[1282] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1283] The system of the present invention combines an emotion engine with a high-performance surveillance camera system that utilizes AI, and is capable of identifying people and objects, detecting abnormal behavior and emergency situations, automatically issuing alarms, and recognizing the user's emotions and responding based on those emotions. The following describes in detail the embodiments of the present invention.

[1284] System Configuration

[1285] The system includes the following main components:

[1286] 1. Terminal (surveillance camera)

[1287] 2. Server

[1288] 3. Emotion Engine

[1289] 4. User devices (smartphones and PCs)

[1290] 5. Related agencies (police, fire department, etc.)

[1291] Acquiring and transmitting camera images

[1292] The device (surveillance camera) captures real-time video of a specified area. The video is divided into frames and temporarily stored in a buffer. The video data in the buffer is compressed at regular intervals and sent to the server.

[1293] Receiving and analyzing video data

[1294] The server receives the video data sent from the device. The received data is decompressed and immediately input into the AI ​​model. This model identifies people and objects, and detects abnormal behavior and emergencies. The detected data is recorded in a database.

[1295] Detecting abnormal behavior and issuing alerts

[1296] The server captures and flags any abnormal behavior or emergency situations, capturing their timestamp and location information. The video clip, timestamp, and location information are saved and generated as an alert message, which is then sent to the appropriate authorities in the form of a push notification.

[1297] Introducing the Emotion Engine

[1298] The emotion engine uses the captured video data and abnormal behavior analysis data to recognize emotions from the user's facial expressions and voice. This emotion data is used to evaluate abnormal behavior and optimize responses.

[1299] Optimizing responses based on perceived emotions

[1300] The server uses the emotion data obtained from the emotion engine to optimize the content of the alert and emergency response. For example, if a user is in a panic state, the server sends an alert that increases the level of urgency to the relevant authorities.

[1301] User Notifications

[1302] User devices (smartphones and PCs) receive notifications when abnormal behavior or emergencies occur, including a detailed description of the event, a link to a video clip, and appropriate countermeasures based on the user's feelings.

[1303] Cooperation with related organizations

[1304] The server waits for confirmation of the alert sent to the relevant authorities, and once confirmation is received, it provides the next important update (e.g., changes in the situation on the ground, additional video clips, etc.) in real time, allowing the relevant authorities to respond quickly and accurately.

[1305] Big Data Analysis and Prediction

[1306] The server uses the accumulated video data and emotion data for big data analysis, training predictive models to forecast crimes and emergencies, and supporting optimal resource allocation and countermeasure planning based on the prediction results.

[1307] Specific examples

[1308] Example 1: Incident detection on public transport

[1309] The device (surveillance camera) continues to monitor the video near the ticket gate of the station. Among the people passing by, it detects one person holding a knife.

[1310] The server detects this abnormal behavior and uses an emotion engine to analyze the emotions of the person wielding the knife and those around them. If emotions of panic or fear are recognized, a high-level alert is generated and sent to the relevant authorities.

[1311] The relevant authorities (police) receive a notification and dispatch the nearest police officer to the scene. The server continues to monitor the footage from the camera and provides additional information to the relevant authorities.

[1312] The user (station staff) will receive a notification on their smartphone, and will be able to immediately grasp the situation and prepare a response.

[1313] Example 2: Suspicious behavior detection in a commercial area

[1314] The device (surveillance camera) monitors the video at the entrance of the shopping mall and detects the behavior of a person stealing goods from multiple stores.

[1315] The server detects this abnormal behavior and uses an emotion engine to analyze the emotions of the thief and those around them. If emotions of anxiety or confusion are recognized, an alert is generated suggesting appropriate responses and sent to the relevant authorities.

[1316] The relevant authorities (police) receive the notification and immediately dispatch the nearest police officers to the scene. The server updates the relevant authorities as the situation changes.

[1317] The user (store staff) will receive a notification on their smartphone, immediately check the security camera footage, and take appropriate action while waiting for the police to arrive.

[1318] In this way, by combining an emotion engine with an AI-powered surveillance camera system, abnormal behavior can be detected in real time, strengthening collaboration with users and relevant agencies, and enabling swift and accurate responses. By utilizing big data analysis and predictive models, crime prevention and efficient resource allocation become possible.

[1319] The processing flow will be explained below.

[1320] Step 1:

[1321] The device (surveillance camera) captures real-time video of the specified area. The video is divided into frames and temporarily stored in a buffer.

[1322] Step 2:

[1323] The terminal (surveillance camera) compresses the video data in the buffer at regular intervals (e.g., every second) and sends it to a server via the Internet.

[1324] Step 3:

[1325] The server receives the video data sent from the device, decompresses it, and immediately inputs it into the AI ​​model.

[1326] Step 4:

[1327] The server analyzes the transmitted video data and identifies people and objects using an AI model that is optimized based on pre-trained data.

[1328] Step 5:

[1329] Based on the analysis results, the server tracks the movements of people and objects in the video and uses an abnormal behavior detection algorithm to detect abnormal behavior such as falls, fights, or the possession of knives.

[1330] Step 6:

[1331] When the server detects abnormal or emergency activity, it clips the frame of data and flags it as abnormal activity. The video clip, timestamp, and location information are saved.

[1332] Step 7:

[1333] The server uses an emotion engine to analyze the user's facial expressions and voice based on the analysis results of the video data and abnormal behavior, and recognizes emotions.

[1334] Step 8:

[1335] The server then optimizes the content of the alert and emergency response based on the recognized emotion. For example, if a user is in a panic state, it will generate a high-urgency alert and send it to the relevant authorities.

[1336] Step 9:

[1337] The server sends an alert to the relevant authorities (police, fire department) in the form of a push notification. The alert message includes details of the incident, a video clip, a timestamp, and location information.

[1338] Step 10:

[1339] The relevant authorities (police and fire departments) receive an alert notification and begin preparing to respond to the incident, dispatching the nearest police officers and firefighters to the scene.

[1340] Step 11:

[1341] The server monitors the response status of relevant agencies and continues to transmit video data in real time if the situation at the scene changes, providing the latest information.

[1342] Step 12:

[1343] The server sends notifications of emergencies or abnormal behavior to specific user devices (e.g., commercial facility staff, security personnel), including detailed descriptions, links to video clips, and appropriate countermeasures based on the recognized emotions.

[1344] Step 13:

[1345] Users (such as commercial facility staff) receive notifications on their smartphones or PCs, immediately check the situation, and take action to ensure the safety of their surroundings and prevent damage, if necessary.

[1346] Step 14:

[1347] The server stores the accumulated video data and emotion data and uses it for big data analysis, using predictive models to forecast future crimes and emergencies.

[1348] Step 15:

[1349] The server generates detailed reports based on the analysis results to help optimize resource allocation and plan crime prevention activities, and the generated reports enable efficient countermeasures to be taken.

[1350] This enables the system to utilize AI for real-time monitoring and immediate response, and employs an emotion engine to take appropriate action in the event of abnormal behavior or an emergency, significantly improving safety in cities, commercial districts, public transportation, and communities.

[1351] Example 2

[1352] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1353] Current surveillance camera systems are limited to simply monitoring video footage and detecting abnormal behavior, and lack the ability to accurately assess users' emotions and situations and respond quickly and appropriately. Furthermore, their functionality for coordinating with relevant agencies and preventing abnormal behavior is limited. This leads to problems such as delayed initial responses to emergencies and difficulty allocating appropriate resources.

[1354] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1355] In this invention, the server includes a means for acquiring camera footage, a means for analyzing the acquired camera footage in real time and identifying people or objects, a means for detecting abnormal behavior or emergencies, a means for automatically issuing an alert based on the detected abnormal behavior or emergency, a means for coordinating with relevant organizations in real time, and a means for recognizing emotions from the user's facial expressions and voice and optimizing responses. This integrates the process from detecting abnormal behavior to emergency response, enabling flexible responses based on the user's emotions. Furthermore, the combined use of big data analysis and predictive models enables more accurate prediction and prevention of abnormal behavior and emergencies, enabling optimal resource allocation and rapid response.

[1356] "Camera footage" refers to visual information obtained from a surveillance camera.

[1357] "Video data" refers to data that represents camera images in digital format.

[1358] "Real-time analysis" refers to processing and analyzing data as soon as it is received.

[1359] "Person or object identification" refers to analyzing surveillance camera footage and identifying people or objects present in the footage.

[1360] "Abnormal behavior" refers to behavior that deviates from general patterns of behavior and is considered dangerous or suspicious.

[1361] An "emergency" is a sudden situation or event that requires immediate action.

[1362] "Alert" refers to a warning or notification issued when abnormal behavior or an emergency is detected.

[1363] "Related agencies" refers to external organizations (e.g., police, fire department, etc.) that respond to abnormal behavior or emergencies.

[1364] "Facial expressions" refer to the movements and changes of facial muscles that express a person's emotions and psychological state.

[1365] "Audio" refers to data related to human voices and sounds.

[1366] "Emotion recognition" refers to identifying a user's emotions and mental state from their facial expressions and voice.

[1367] "Optimizing responses" refers to selecting and implementing the most appropriate and effective response measures based on emotional data and the situation.

[1368] "Big data analytics" refers to the process of analyzing large amounts of data and extracting meaningful information and patterns.

[1369] "Predictive model" refers to a mathematical or statistical model used to forecast future events or trends based on historical data.

[1370] The system of the present invention combines an emotion engine with a high-performance surveillance camera system that utilizes AI, and is capable of identifying people and objects, detecting abnormal behavior and emergency situations, automatically issuing alarms, and recognizing the user's emotions and responding based on those emotions. The following are modes for implementing the present invention.

[1371] System Configuration

[1372] The system includes the following main components:

[1373] 1. Terminal (surveillance camera)

[1374] 2. Server

[1375] 3. Emotion Engine

[1376] 4. User devices (smartphones and PCs)

[1377] 5. Related agencies (police, fire department, etc.)

[1378] Acquiring and transmitting camera images

[1379] The device (surveillance camera) captures real-time video of a specified area. For example, this applies to surveillance cameras installed in surveillance systems for public transportation or commercial districts. The video is divided into frames and temporarily stored in a buffer. The video data is periodically compressed using a compression technology such as H.264 and sent to a server using a secure protocol such as HTTPS.

[1380] Receiving and analyzing video data

[1381] The server receives the compressed video data sent from the device. The received data is decompressed and input into a generative AI model (e.g., YOLOv4 or OpenPose). This AI model identifies people and objects and detects abnormal behavior and emergencies. For example, it can detect dangerous behavior such as someone carrying a knife on public transportation.

[1382] Detecting abnormal behavior and issuing alerts

[1383] The server then assigns a timestamp and location information to any abnormal behavior or emergency detected by the AI ​​model. Based on this, an alert message is generated and sent to the relevant authorities in the form of a push notification, along with a detailed description of the incident and a link to a video clip. For example, police and fire departments are notified immediately.

[1384] Introducing an emotion engine and optimizing responses

[1385] The emotion engine analyzes the user's facial expressions and voice from the acquired video data and abnormal behavior analysis data to recognize their emotions. The server can use this emotion data to optimize the content of warnings and emergency responses. For example, if a user is in a panic state, it can send a high-level alert to the relevant authorities.

[1386] User notification and collaboration with relevant organizations

[1387] User devices (smartphones and PCs) receive notifications when abnormal behavior or emergencies occur. The notifications include a detailed description of the event, a link to a video clip, and appropriate countermeasures based on emotion data, allowing users to respond quickly.

[1388] Big Data Analysis and Prediction

[1389] The server then performs big data analysis on the accumulated video and emotion data, which can then be used to train predictive models to forecast future abnormal behavior and emergencies. The prediction results can then be used to optimize resource allocation and support countermeasure planning.

[1390] Resulting System Behavior

[1391] This system analyzes real-time footage from surveillance cameras, detects abnormal behavior and emergencies, and immediately coordinates with relevant agencies to take appropriate action. Furthermore, it has the ability to analyze user emotions and respond appropriately based on those emotions, enabling rapid and effective response to emergencies. Big data analysis also improves prediction accuracy, enabling advance countermeasures and optimal resource allocation.

[1392] Examples of prompt statements

[1393] "Please explain in detail the process of the system that detects abnormal behavior at station ticket gates and also analyzes the emotions of those around you."

[1394] "Please explain a system that detects suspicious behavior in a shopping mall and performs sentiment analysis based on that behavior."

[1395] In this way, the system of the present invention enables real-time detection of abnormal behavior, prompt notification to users and relevant organizations, and optimal response based on emotional data.

[1396] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1397] Step 1:

[1398] The terminal (surveillance camera) acquires real-time video of the designated area.

[1399] Input: Video of the specified area

[1400] Processing: Capture video in frames (e.g. 30 frames per second) and save as frame data

[1401] Output: Frame data

[1402] Step 2:

[1403] The terminal (surveillance camera) stores the acquired frame data in a buffer, compresses it at regular intervals, and sends it to the server.

[1404] Input: Frame data

[1405] Processing: Frame data is temporarily stored in a buffer and compressed using a compression algorithm such as H.264

[1406] Output: Compressed video data is sent to the server

[1407] Step 3:

[1408] The server receives the compressed video data sent from the terminal and decompresses it.

[1409] Input: Compressed video data

[1410] Processing: Decompresses received data and restores it as frame data

[1411] Output: Decompressed frame data

[1412] Step 4:

[1413] The server inputs the decompressed frame data into a generative AI model to identify people or objects.

[1414] Input: Decompressed frame data

[1415] Processing: Input frame data into a generative AI model (e.g., YOLOv4, OpenPose, etc.) to identify people and objects.

[1416] Output: Data of identified people or objects (location, label, etc.)

[1417] Step 5:

[1418] The server analyzes the data of the identified person or object to detect abnormal behavior or emergency situations.

[1419] Input: Identified person or object data

[1420] Processing: Uses analytical algorithms to detect abnormal behavior and emergencies

[1421] Output: Data on detected abnormal behavior or emergency (e.g., timestamp, location, behavior type, etc.)

[1422] Step 6:

[1423] The server issues alerts based on detected abnormal behavior or emergency situations and sends push notifications to relevant authorities.

[1424] Input: Data on abnormal behavior and emergencies

[1425] Processing: Generate an alert message and send it to the relevant authorities in the form of a push notification.

[1426] Output: Alert message to relevant organizations

[1427] Step 7:

[1428] The emotion engine uses the acquired video data and abnormal behavior analysis data to recognize emotions from the user's facial expressions and voice.

[1429] Input: Video data, abnormal behavior analysis data

[1430] Processing: Apply emotion recognition algorithms to identify the user's emotions

[1431] Output: Recognized emotion data (e.g., panic, fear, anxiety, etc.)

[1432] Step 8:

[1433] The server uses the recognized emotion data to optimize alert content and emergency response.

[1434] Input: Emotion data

[1435] Processing: Adjust the content and urgency of the warning message and optimize the necessary response measures.

[1436] Output: Optimized warning messages and countermeasures

[1437] Step 9:

[1438] User devices (smartphones and PCs) receive notifications when abnormal behavior or emergencies occur.

[1439] Input: Optimized warning message

[1440] Action: Display a notification and inform the user of the situation

[1441] Output: A message to inform the user

[1442] Step 10:

[1443] The server waits for confirmation of receipt of alerts sent to relevant authorities and provides important updates in real time.

[1444] Input: Acknowledgments from relevant agencies and new video data

[1445] Processing: Get acknowledgements and real-time updates on important information like changing conditions on the ground and additional footage clips

[1446] Output: Update information for relevant organizations

[1447] Step 11:

[1448] The server uses the accumulated video data and emotion data for big data analysis.

[1449] Input: Accumulated video data, emotion data

[1450] Processing: Data is stored in a data warehouse and analyzed using data mining techniques

[1451] Output: Analysis results and training data for predictive models

[1452] Step 12:

[1453] The server uses predictive models to forecast crimes and emergencies and help plan countermeasures.

[1454] Input: Prediction model, analysis results

[1455] Processing: Applying predictive algorithms to predict future abnormal behavior and emergencies

[1456] Output: Prediction results and supporting information for action plans

[1457] (Application example 2)

[1458] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1459] Conventional surveillance systems have the ability to identify people and objects and detect abnormal behavior and emergencies, but they are unable to recognize user emotions and provide responses based on those emotions. Furthermore, responses after detecting abnormal behavior are limited, often resulting in failure to contact relevant authorities or propose appropriate countermeasures. To address these issues, the present invention aims to provide a system that adds emotion recognition functionality and optimizes responses after detecting abnormal behavior.

[1460] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for acquiring camera images for identifying people or objects, means for analyzing camera image data and identifying people or objects, means for detecting abnormal behavior or emergency situations, and means for recognizing user emotions and optimizing responses based on the emotions. This makes it possible to take appropriate action based not only on identification but also on emotion data when abnormal behavior or an emergency occurs.

[1461] "Person or object identification" refers to technology that analyzes camera video data to identify people or objects.

[1462] "Unusual behavior or emergency" means any behavior that deviates from normal patterns of behavior or any situation that poses an immediate danger to life or property.

[1463] "Automatic alerting means" means a technology that automatically issues an alert in accordance with pre-defined conditions based on detected abnormal behavior or emergency situations.

[1464] "Means for recognizing the user's emotions and optimizing the response based on that" refers to technology that analyzes the user's emotions using acquired video and audio data and provides the optimal response based on the results.

[1465] "Means of coordinating with relevant agencies in real time" refers to technology that allows for instant communication with relevant agencies such as the police and fire department, providing necessary information, and maintaining a cooperative relationship.

[1466] "Big data analytics and predictive models" are techniques for analyzing large amounts of data to derive specific patterns and trends, which can be used to predict abnormal behavior and emergencies.

[1467] "Means for providing video clips, timestamps, and location information" refers to technology that, when issuing an alert, adds detailed information such as specific video data, the time of occurrence, and location.

[1468] "Means for proposing appropriate countermeasures" refers to technology that presents optimal and effective countermeasures based on detected abnormal behavior and user emotional data.

[1469] System Overview

[1470] The present invention is a system that identifies people and objects, detects abnormal behavior and emergencies, and automatically issues an alarm. Furthermore, it recognizes the user's emotions and provides an optimal response based on those emotions, making the response after abnormal behavior is detected more effective.

[1471] System Components

[1472] The system includes the following major components:

[1473] 1. Surveillance cameras

[1474] 2. Server

[1475] 3. Emotion Recognition Engine

[1476] 4. User devices (smartphones and PCs)

[1477] 5. Related agencies (police, fire department, etc.)

[1478] System Operation

[1479] Acquiring and transmitting camera images

[1480] The surveillance camera captures real-time video of a specified area. The video is divided into frames and temporarily stored in a buffer. The video data in the buffer is compressed at regular intervals and sent to the server.

[1481] Receiving and analyzing video data

[1482] The server receives video data transmitted from the surveillance cameras. The received data is decompressed and immediately input into a generative AI model. This model identifies people and objects, and detects abnormal behavior and emergencies. The detected data is then recorded in a database.

[1483] Detecting abnormal behavior and issuing alerts

[1484] The server captures and flags any abnormal behavior or emergency situations, capturing their timestamp and location information. The video clip, timestamp, and location information are saved, and an alert message is generated and sent to the appropriate authorities in the form of a push notification.

[1485] Using the Emotion Recognition Engine

[1486] The server uses the acquired video data and abnormal behavior analysis data to recognize emotions from the user's facial expressions and voice using an emotion recognition engine. This emotion data is used to evaluate abnormal behavior and optimize responses.

[1487] Optimizing responses based on perceived emotions

[1488] The server uses the emotion data obtained from the emotion recognition engine to optimize the content of the alert and emergency response, for example, sending an alert to the relevant authorities that increases the level of urgency if the user is in a panic.

[1489] User notification and response

[1490] The user device receives a notification when abnormal behavior or an emergency occurs. The notification includes a detailed description of the event, a link to a video clip, and appropriate countermeasures based on the user's emotions. The notification arrives on the user's smartphone, allowing the user to immediately understand the situation and prepare a response.

[1491] Hardware and software used

[1492] Hardware:

[1493] Surveillance cameras (IP cameras, etc.)

[1494] Smartphone or PC

[1495] software:

[1496] OpenCV: Camera image acquisition and analysis

[1497] Keras: Loading and predicting emotion recognition models

[1498] requests: Server alert notification

[1499] Specific processing examples

[1500] Example 1: Suspicious behavior detection in a shopping mall

[1501] 1. A surveillance camera monitors footage at the entrance to a shopping mall and detects an individual stealing items from multiple stores.

[1502] 2. The server detects this abnormal behavior and uses an emotion recognition engine to analyze the emotions of the thief and those around them. If emotions of anxiety or confusion are recognized, an alert is generated suggesting appropriate responses and sent to the relevant authorities.

[1503] 3. The relevant authorities (police) receive the notification and immediately dispatch the nearest police officers to the scene. The server updates the relevant authorities with information as the situation changes.

[1504] 4. The user (store staff) receives a notification on their smartphone, immediately checks the security camera footage, and takes appropriate action while waiting for the police to arrive.

[1505] Prompt Sentence Examples

[1506] "Detect anomalous behavior in this camera footage and recognize the emotions of people around it. If anomalous behavior is detected, generate an alert message suggesting appropriate countermeasures based on the emotional data."

[1507] The above is an embodiment of the present invention. The present invention makes it possible to detect abnormal behavior in real time, strengthen cooperation with users and related organizations, and realize a prompt and accurate response.

[1508] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1509] Step 1:

[1510] The surveillance camera captures real-time video of a specified area. The input is the video data from the surveillance camera, and the output is the video frame stored in the buffer. This frame is temporarily stored for further processing.

[1511] Step 2:

[1512] The server compresses the video data in the buffer at regular intervals and sends it to the server. The input is the video frames stored in the buffer, and the output is the compressed video data. Data compression enables efficient use of network bandwidth.

[1513] Step 3:

[1514] The server receives the compressed video data and decompresses it. The input is the compressed video data, and the output is the decompressed video data. The decompressed data is immediately input into the generative AI model.

[1515] Step 4:

[1516] The server uses a generative AI model to identify people and objects. The input is the decompressed video data, and the output is the information of the identified people and objects. The generative AI model identifies people and objects based on video analysis.

[1517] Step 5:

[1518] The server further analyzes the information of the identified people and objects to detect abnormal behavior or emergency situations. The input is the information of the identified people and objects, and the output is a flag indicating abnormal behavior or emergency situation. An algorithm is used to detect patterns of abnormal behavior or emergency situations.

[1519] Step 6:

[1520] When an abnormal behavior or emergency is detected, the server acquires the timestamp and location information and sets a flag. The input is the flag for abnormal behavior detection or emergency detection, and the output is the timestamp and location information.

[1521] Step 7:

[1522] The server generates an alert message based on the video clip, timestamp, and location information, and sends it to the relevant organizations. The input is the timestamp, location information, and video clip, and the output is the generated alert message. It is sent to the relevant organizations in the form of a push notification.

[1523] Step 8:

[1524] The server uses an emotion recognition engine to analyze the user's facial expressions and voice based on the acquired video data and recognizes their emotions. The input is video data and the output is emotion data. The emotion recognition engine estimates the user's emotions from their facial expressions and voice.

[1525] Step 9:

[1526] The server optimizes the content of the alert and emergency response using the emotion data obtained from the emotion recognition engine. The input is emotion data, and the output is optimized alert content and emergency response measures. The urgency and content of the response are adjusted based on the emotion data.

[1527] Step 10:

[1528] The user device receives notifications when abnormal behavior or an emergency occurs. The input is the optimized alert content and emergency response measures, and the output is a notification displayed on the user device. The user receives the notification on their smartphone, allowing them to immediately understand the situation and prepare a response.

[1529] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1530] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1531] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1532] [Fourth embodiment]

[1533] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1534] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1535] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1536] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1537] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1538] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1539] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1540] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1541] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1542] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1543] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1544] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1545] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1546] The system of the present invention is a high-performance surveillance camera system that utilizes AI, and can identify people and objects, detect abnormal behavior and emergency situations, and automatically issue an alarm. The following describes in detail an embodiment of the present invention.

[1547] System Configuration

[1548] The system includes the following main components:

[1549] 1. Terminal (surveillance camera)

[1550] 2. Server

[1551] 3. User device (smartphone or PC)

[1552] 4. Related agencies (police, fire department, etc.)

[1553] Acquiring and transmitting camera images

[1554] The devices (surveillance cameras) are placed in cities, commercial districts, public transportation facilities, and communities. These cameras capture images in real time and store them in a buffer frame by frame. The video data in the buffer is compressed and sent to a server at regular intervals.

[1555] Receiving and analyzing video data

[1556] The server receives the video data sent from the device. The received data is decompressed and input into an AI model. This model identifies people and objects, and detects abnormal behavior and emergencies. For example, it can detect situations where a person falls or a fight breaks out. The detected data is recorded in a database.

[1557] Detecting abnormal behavior and issuing alerts

[1558] When an abnormal behavior or emergency situation is detected, the server acquires the timestamp and location information and prepares it for the emergency notification system. The alert is generated as an alert message containing the video clip, timestamp, and location information, and is sent to the relevant authorities in the form of a push notification.

[1559] Cooperation with related organizations

[1560] The server waits for confirmation of the alert sent to the relevant authorities, and once confirmation is received, it provides the next important update (e.g., changes in the situation on the ground, additional video clips, etc.) in real time, allowing the relevant authorities to respond quickly and accurately.

[1561] User Notifications

[1562] The server sends notifications to specific user devices when an emergency or abnormal behavior occurs, including a detailed description of the event and a link to a video clip, allowing users to take immediate action.

[1563] Big Data Analysis and Prediction

[1564] The server performs big data analysis using accumulated video data and abnormal behavior detection logs. It trains predictive models to forecast crimes and emergencies. Based on the prediction results, it helps optimize resource allocation and plan countermeasures. This improves the efficiency of crime prevention and emergency response.

[1565] Specific examples

[1566] Example 1: Incident detection on public transport

[1567] The device (surveillance camera) continues to monitor the video near the ticket gate of the station. Among the people passing by, it detects one person holding a knife.

[1568] The server detects this anomalous behavior and sends a notification to the emergency notification system, which includes the video clip, a timestamp, and location information.

[1569] The relevant authorities (police) receive a notification and dispatch the nearest police officer to the scene. The server continues to monitor the footage from the camera and provides additional information to the relevant authorities.

[1570] The user (station staff) will receive a notification on their smartphone, and will be able to immediately grasp the situation and prepare a response.

[1571] Example 2: Suspicious behavior detection in a commercial area

[1572] The device (surveillance camera) monitors the video at the entrance of the shopping mall and detects the behavior of a person stealing goods from multiple stores.

[1573] The server detects this unusual behavior and sends an alert to the police, along with video clips and location information, for rapid response.

[1574] The relevant authorities (police) receive the notification and immediately dispatch the nearest police officers to the scene. The server updates the relevant authorities as the situation changes.

[1575] The user (store staff) will receive a notification on their smartphone, immediately check the security camera footage, and take appropriate action while waiting for the police to arrive.

[1576] This allows the AI-powered surveillance camera system to detect abnormal behavior in real time, strengthen cooperation with relevant agencies, and enable rapid response. It also utilizes big data analytics and predictive models to prevent crime and efficiently allocate resources.

[1577] The processing flow will be explained below.

[1578] Step 1:

[1579] The device (surveillance camera) captures real-time video of the specified area. The video is divided into frames and temporarily stored in a buffer.

[1580] Step 2:

[1581] The terminal (surveillance camera) compresses the video data in the buffer at regular intervals (e.g., every second) and sends it to a server via the Internet.

[1582] Step 3:

[1583] The server receives the video data sent from the device, decompresses it, and immediately inputs it into the AI ​​model.

[1584] Step 4:

[1585] The server analyzes the transmitted video data and identifies people and objects using an AI model that is optimized based on pre-trained data.

[1586] Step 5:

[1587] Based on the analysis results, the server tracks the movements of people and objects in the video and uses an abnormal behavior detection algorithm to detect abnormal behavior such as falls, fights, or the possession of knives.

[1588] Step 6:

[1589] When the server detects abnormal or emergency activity, it clips the frame of data and flags it as abnormal activity. The video clip, timestamp, and location information are saved.

[1590] Step 7:

[1591] If any unusual behavior or emergency is flagged, the server immediately generates an alert message, which includes details of the incident, video clips, timestamps, and location information.

[1592] Step 8:

[1593] The server sends an alert to the relevant authorities (police and fire departments) in the form of a push notification. After sending the alert, it waits for confirmation of receipt of the alert, and once confirmation is received, it proceeds to the next step.

[1594] Step 9:

[1595] The relevant authorities (police and fire departments) receive an alert notification and begin preparing to respond at the scene, dispatching the nearest police officers and firefighters to the scene.

[1596] Step 10:

[1597] The server monitors the response status of relevant agencies and continues to transmit video data in real time if the situation at the scene changes, providing the latest information.

[1598] Step 11:

[1599] The server sends notifications of emergencies or unusual behavior to specific user devices (e.g., commercial facility staff, security personnel), including detailed descriptions and links to video clips.

[1600] Step 12:

[1601] Users (such as commercial facility staff) receive notifications on their smartphones or PCs, immediately check the situation, and take action to ensure the safety of their surroundings and prevent damage, if necessary.

[1602] Step 13:

[1603] The server stores the accumulated video data and abnormal behavior detection logs, and uses them for big data analysis, using predictive models to forecast future crimes and emergencies.

[1604] Step 14:

[1605] The server generates detailed reports based on the analysis results to help optimize resource allocation and plan crime prevention activities, and the generated reports enable efficient countermeasures to be taken.

[1606] This enables the system to leverage AI for real-time monitoring and immediate response, significantly improving safety in cities, commercial districts, public transportation, and communities.

[1607] Example 1

[1608] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1609] Conventional surveillance camera systems have difficulty detecting abnormal behavior and emergencies in real time and providing prompt notification. Furthermore, due to insufficient collaboration with relevant agencies and insufficient predictive capabilities based on big data analysis, there were issues with the efficiency of crime prevention and emergency response. Furthermore, the notification functionality was limited, preventing users from taking immediate action.

[1610] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1611] In this invention, the server includes a device for acquiring camera footage, a device for compressing the acquired video data and sending it to the server, a device for decompressing the data received by the server and inputting it into an AI model to identify people or objects, a device for detecting abnormal behavior or emergencies, a device for recording information on the detected abnormal behavior or emergency in a database and generating an alert, a device for coordinating with relevant organizations in real time, and a device for sending notifications of emergency situations or abnormal behavior to specific user devices. This enables real-time detection of abnormal behavior or emergencies and rapid notification, strengthens cooperation with relevant organizations, and enables crime prevention and efficient resource allocation through predictions using big data analysis.

[1612] A "device for acquiring camera images" is a device that has the function of capturing images in real time in the environment in which it is installed and storing the image data.

[1613] The "device that compresses video data and transmits it to a server" is a device that has the function of compressing acquired video data and transmitting it to a server in an efficient format.

[1614] "Device that decompresses received data and inputs it into an AI model" refers to a device that has the function of decompressing compressed data received by the server and inputting it into an AI model.

[1615] A "device for identifying people or objects" is a device that has the function of identifying people or objects in video data using an AI model.

[1616] The "device for detecting abnormal behavior or emergency situations" is a device that has the function of detecting abnormal behavior or emergency situations that exceed a predetermined standard based on identified data.

[1617] An "alert generating device" is a device that has the function of generating and notifying an alert based on information on detected abnormal behavior or emergency situations.

[1618] The "device that cooperates with relevant organizations in real time" is a device that has the function of notifying relevant organizations of generated alerts in real time and ensuring cooperation.

[1619] A "device that transmits a notification to a specific user device" is a device that has a function of transmitting a notification to a specific user device when abnormal behavior or an emergency occurs.

[1620] A "device that uses big data analysis and predictive models" is a device that has the function of predicting future abnormal behavior or emergency situations using big data analysis and predictive models based on accumulated video data and abnormal behavior detection logs.

[1621] A "database recording device" is a device that has the function of storing information on detected abnormal behavior or emergency situations in a database.

[1622] MODE FOR CARRYING OUT THE INVENTION

[1623] The present invention relates to a high-performance surveillance camera system that utilizes AI. Hereinafter, an embodiment of the present invention will be described in detail.

[1624] System Configuration

[1625] The system mainly consists of the following components:

[1626] 1. Terminal (surveillance camera)

[1627] 2. Server

[1628] 3. User device (smartphone or PC)

[1629] 4. Related agencies (police, fire department, etc.)

[1630] Acquiring camera footage

[1631] The devices (surveillance cameras) are placed in cities, commercial districts, public transportation facilities, and communities. These cameras capture video in real time and store it in a buffer frame by frame. This stored video data is recompressed at regular intervals and sent to a server.

[1632] Video data compression and transmission

[1633] The device compresses the captured video data using a compression method such as H.264. The compressed data is then encrypted and securely sent to the server using the HTTPS protocol.

[1634] Receiving and decompressing video data

[1635] The server receives the compressed data sent from the device, stores it temporarily, and then decompresses it using decompression software such as FFmpeg to restore the original video.

[1636] Identification and detection with AI models

[1637] The server uses a deep learning framework (e.g., TensorFlow or PyTorch) to input the decompressed video data into an AI model, which then identifies people and objects and detects abnormal behavior and emergencies.

[1638] Recording abnormal behavior and generating alerts

[1639] The server records the detected abnormal behavior and emergency information in a database (e.g., MySQL, PostgreSQL), and simultaneously generates an alert message and sends it to the emergency notification system.

[1640] Sending and acknowledging alerts

[1641] The server sends the generated alert message to the relevant organizations in the form of a push notification. After sending, it waits for an ACK (acknowledgment) and, once it receives the ACK, provides additional information in real time.

[1642] User Notifications

[1643] The server notifies specific user devices of emergencies or abnormal behavior, including a detailed description of the event and a link to a video clip, allowing users to take immediate action.

[1644] Big Data Analysis and Prediction

[1645] The server uses the accumulated video data and abnormal behavior detection logs to perform big data analysis and predictive models. It processes the data using big data processing tools such as Hadoop and Spark to train predictive models. This makes it possible to predict crimes and emergencies, and helps with optimal resource allocation and countermeasure planning.

[1646] Specific examples

[1647] Example 1: Incident detection on public transport

[1648] The device (surveillance camera) continues to monitor the video near the ticket gate of the station. Among the people passing by, it detects one person holding a knife.

[1649] The server detects this unusual behavior and sends a notification to the emergency notification system, which includes a video clip, a timestamp, and location information.

[1650] The relevant authorities (police) are notified and dispatch the nearest police officers to the scene. The server continues to monitor the footage from the camera and provides additional information to the relevant authorities.

[1651] The user (station staff) will receive a notification on their smartphone, and will be able to immediately grasp the situation and prepare a response.

[1652] Prompt Sentence Examples

[1653] "Please describe an AI model that instantly detects a person with a knife at a station ticket gate and sends a notification to the appropriate authorities via an emergency notification system."

[1654] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1655] Step 1:

[1656] The terminals (surveillance cameras) capture images in real time in the areas where they are installed.

[1657] Input: Video data from a surveillance camera sensor

[1658] Data processing: Capture video frame by frame

[1659] Output: Video data divided into frames is stored in a buffer

[1660] Specific operation: The surveillance camera captures 30 frames of video per second and stores them in buffer memory.

[1661] Step 2:

[1662] The terminal compresses the video data in the buffer at regular intervals.

[1663] Input: Buffered raw video data

[1664] Data processing: Compresses video using compression algorithms such as H.264

[1665] Output: Compressed video data

[1666] Specific operation: Every 10 seconds, the device compresses the video data stored in the buffer and stores it in the output buffer.

[1667] Step 3:

[1668] The terminal transmits the compressed video data to the server.

[1669] Input: Compressed video data

[1670] Data processing: Encrypting data and preparing it for transmission using the HTTPS protocol

[1671] Output: Secure compressed video data sent to the server

[1672] Specific operation: The terminal encrypts the compressed video data using HTTPS and sends it to the server.

[1673] Step 4:

[1674] The server receives the compressed data sent from the terminal.

[1675] Input: Compressed and encrypted video data

[1676] Data processing: Decrypt the received data and store it in temporary storage

[1677] Output: Compressed data saved to storage

[1678] Specific operation: The server receives data sent via HTTPS, decrypts it, and stores it in storage.

[1679] Step 5:

[1680] The server decompresses the received data.

[1681] Input: Compressed video data

[1682] Data processing: Decompress the compressed data using decompression software such as FFmpeg.

[1683] Output: Original video data

[1684] Specific operation: The server uses FFmpeg to decompress the compressed data and return it to the original video frames.

[1685] Step 6:

[1686] The server inputs the decompressed video data into the AI ​​model.

[1687] Input: Decompressed video data

[1688] Data processing: Analyze video using deep learning frameworks such as TensorFlow and PyTorch

[1689] Output: Recognition results (identification of people and objects, detection of abnormal behavior)

[1690] How it works: The AI ​​model inputs video data frame by frame to detect abnormal behavior or emergencies.

[1691] Step 7:

[1692] The server records any abnormal behavior or emergency situations and generates alerts.

[1693] Input: Recognition results by AI model

[1694] Data processing: Recording abnormal behavior and emergency information in a database and generating alerts

[1695] Output: Data recorded in the database, generated alert messages

[1696] Specific operation: The server records abnormal behavior information in MySQL or PostgreSQL and generates an alert.

[1697] Step 8:

[1698] The server sends an alert message to the appropriate authority and waits for an acknowledgement.

[1699] Input: The generated alert message

[1700] Data processing: Generate and send a push notification containing the alert message

[1701] Output: Notification sent to the relevant authorities, acknowledgement (ACK)

[1702] Specific operation: The server sends a notification using REST API or WebSocket and waits for confirmation of receipt.

[1703] Step 9:

[1704] The server sends a notification to the specific user terminal.

[1705] Input: Information about abnormal behavior or emergency situations

[1706] Data processing: Generate notification messages and send them to the user terminal

[1707] Output: Notification sent to the user's device

[1708] Specific operation: The server uses a push notification service (e.g., Firebase Cloud Messaging) to send a notification to the user device.

[1709] Step 10:

[1710] The server performs big data analysis and predictions.

[1711] Input: Accumulated video data and abnormal behavior detection logs

[1712] Data processing: Analyze data and train predictive models using Hadoop or Spark

[1713] Output: Prediction results, optimal resource allocation proposals

[1714] Specific operation: The server processes the data using big data tools and trains a predictive model to predict future abnormal behavior.

[1715] (Application example 1)

[1716] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1717] In modern society, surveillance systems are important for ensuring public safety. However, current surveillance systems lack the ability to detect abnormal behavior and emergencies in real time, making it difficult for security guards and relevant agencies to respond immediately. Furthermore, they lack the functionality to quickly detect abnormalities on-site using mobile devices and respond appropriately, resulting in a decrease in the efficiency of security services.

[1718] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1719] In this invention, the server includes means for acquiring camera footage for identifying people or objects, means for analyzing camera footage data and identifying people or objects, means for detecting abnormal behavior or emergencies, means for automatically issuing an alarm based on the detected abnormal behavior or emergency, means for coordinating with relevant organizations in real time, means for displaying an alarm on a display device mounted on a mobile terminal, means for transmitting video data from the mobile terminal to the server, and means for transmitting analyzed data and alarm information to relevant organizations. This enables security guards to immediately recognize abnormal behavior on-site and respond quickly, strengthening cooperation with relevant organizations and ensuring public safety.

[1720] "Means for acquiring camera footage to identify people or objects" refers to the function of collecting video data using surveillance cameras.

[1721] "Means for analyzing camera video data and identifying people or objects" refers to algorithms or software for analyzing collected video data and identifying people or objects.

[1722] "Means for detecting abnormal behavior or emergency situations" refers to the ability to use AI or machine learning models to detect behavior that deviates from normal behavior or emergency situations in real time.

[1723] "Means for automatically issuing an alert based on detected abnormal behavior or emergency situations" refers to a function for automatically issuing an alert when abnormal behavior or emergency situations are detected.

[1724] "Means of coordinating with relevant agencies in real time" refers to a system for sharing information in real time with relevant agencies such as the police and fire department and taking joint action.

[1725] "Means for displaying an alert on a display device installed in a mobile terminal" refers to a function for displaying alerts or notification messages in real time on a display device such as a smartphone or smart glasses.

[1726] "Means for transmitting video data from a mobile terminal to a server" refers to a function for uploading video data from a mobile terminal to a server via Internet communication.

[1727] "Means for transmitting analyzed data and warning information to relevant organizations" refers to a system for instantly transmitting AI analysis results and warning notifications to designated relevant organizations.

[1728] MODE FOR CARRYING OUT THE INVENTION

[1729] The present invention provides a security service using a high-performance surveillance camera system and mobile devices that utilize AI. An embodiment of this system is described in detail below.

[1730] System Configuration

[1731] The system consists of the following main components:

[1732] 1. Surveillance camera (terminal)

[1733] 2. Mobile devices (e.g., smart glasses)

[1734] 3. Server

[1735] 4. User devices (smartphones and PCs)

[1736] 5. Related agencies (police, fire department, etc.)

[1737] Acquiring and transmitting camera images

[1738] Surveillance cameras: These cameras are installed in cities, commercial districts, and public facilities and have the ability to capture video in real time. These videos are compressed at regular intervals and sent to a server.

[1739] Receiving and analyzing video data

[1740] Server: Receives and decompresses video data sent from surveillance cameras. The data is then input into an AI model to detect abnormal behavior and emergencies. Examples of AI models used include YOLO and SSD.

[1741] Detection of abnormal behavior: When the server detects abnormal behavior or an emergency, it captures the timestamp and location information and records it in a database.

[1742] Alarm display on mobile device

[1743] Smart glasses: Alerts from the server and notifications about abnormal behavior are displayed on the mobile device in real time to warn the user.

[1744] Video data transmission: On-site video data is uploaded from the smart glasses or smartphone to a server for further analysis.

[1745] Cooperation with related organizations

[1746] Alert notification: The server generates an alert message based on the detected abnormal behavior and sends the analyzed data to the relevant authorities. The notification includes the video clip, timestamp, and location information.

[1747] Big Data Analysis and Prediction

[1748] Big data analysis: The server performs big data analysis using the accumulated video data and abnormal behavior logs, which makes it possible to predict and prevent crimes and emergencies.

[1749] Specific prompt examples

[1750] 1. Example 1: A person is in possession of a knife at a train station.

[1751] 2. Example 2: Theft is detected in a shopping mall.

[1752] Hardware and software used

[1753] Hardware:

[1754] Surveillance cameras: high-resolution cameras (e.g. CCTV cameras)

[1755] Mobile devices: smart glasses (e.g., Google Glass, Vuzix Blade), smartphones

[1756] software:

[1757] AI model: YOLO, SSD

[1758] Server: AWS, Google Cloud

[1759] Notification systems: Firebase, Twilio

[1760] Program processing

[1761] 1. The device (surveillance camera) captures video in real time and sends it to the server at regular intervals.

[1762] 2. The server receives the video data, analyzes it using an AI model, and detects abnormal behavior or emergencies.

[1763] 3. The server generates an alert based on the detected abnormal behavior or emergency situation and displays it on the mobile device (smart glasses, smartphone).

[1764] 4. Send the alert and detailed data to relevant authorities (police, fire department, etc.).

[1765] This will enable security guards to immediately recognize any abnormal behavior on-site and respond quickly, strengthening cooperation with relevant agencies and ensuring public safety.

[1766] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1767] Step 1:

[1768] The terminal (surveillance camera) captures video in real time and temporarily stores the video data in a buffer. At regular intervals, this video data is compressed and sent to the server. The input here is the raw video data captured by the camera, and the output is compressed video data.

[1769] Step 2:

[1770] The server receives the compressed video data sent from the terminal and decompresses it. The decompressed video data is ready for analysis frame by frame. The input here is the compressed video data and the output is the decompressed frame data.

[1771] Step 3:

[1772] The server inputs the decompressed video data into a generative AI model (e.g., YOLO or SSD) to identify people or objects. This identification data is then used to detect abnormal behavior or emergencies. Here, the input is the decompressed frame data, and the output is a list of identified objects.

[1773] Step 4:

[1774] The server analyzes the identified object list to detect anomalous behavior or emergencies. If an anomaly is detected, its timestamp and location information are recorded in a database. The input here is the list of identified objects, and the output is the detection of anomalous behavior or emergencies.

[1775] Step 5:

[1776] The server automatically issues an alert based on the detected abnormal behavior or emergency situation. This alert is sent to the display device of the mobile device (smart glasses or smartphone) and displays a warning to the user. The input here is the abnormal behavior detection result, and the output is the alert message.

[1777] Step 6:

[1778] A mobile device (smart glasses or smartphone) receives an alert message from the server and displays it to the user in real time. The user can then check the details of the alert and take necessary measures. The input here is the alert message, and the output is a notification display to the user.

[1779] Step 7:

[1780] The server sends the alert and detailed data (video clip, timestamp, location information) to the relevant authorities (police, fire department, etc.). The relevant authorities then take prompt action based on this information. The input here is the alert message and detailed data, and the output is a notification to the relevant authorities.

[1781] Step 8:

[1782] The server performs big data analysis and trains a predictive model based on the accumulated video data and abnormal behavior logs. This enables future predictions of crimes and emergencies and allows preventative measures to be planned. The input here is past video data and abnormal behavior logs, and the output is the results of the predictive model.

[1783] By taking the above steps, security guards can immediately recognize any abnormal behavior on-site and respond quickly, strengthening cooperation with relevant agencies and ensuring public safety.

[1784] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1785] The system of the present invention combines an emotion engine with a high-performance surveillance camera system that utilizes AI, and is capable of identifying people and objects, detecting abnormal behavior and emergency situations, automatically issuing alarms, and recognizing the user's emotions and responding based on those emotions. The following describes in detail the embodiments of the present invention.

[1786] System Configuration

[1787] The system includes the following main components:

[1788] 1. Terminal (surveillance camera)

[1789] 2. Server

[1790] 3. Emotion Engine

[1791] 4. User devices (smartphones and PCs)

[1792] 5. Related agencies (police, fire department, etc.)

[1793] Acquiring and transmitting camera images

[1794] The device (surveillance camera) captures real-time video of a specified area. The video is divided into frames and temporarily stored in a buffer. The video data in the buffer is compressed at regular intervals and sent to the server.

[1795] Receiving and analyzing video data

[1796] The server receives the video data sent from the device. The received data is decompressed and immediately input into the AI ​​model. This model identifies people and objects, and detects abnormal behavior and emergencies. The detected data is recorded in a database.

[1797] Detecting abnormal behavior and issuing alerts

[1798] The server captures and flags any abnormal behavior or emergency situations, capturing their timestamp and location information. The video clip, timestamp, and location information are saved and generated as an alert message, which is then sent to the appropriate authorities in the form of a push notification.

[1799] Introducing the Emotion Engine

[1800] The emotion engine uses the captured video data and abnormal behavior analysis data to recognize emotions from the user's facial expressions and voice. This emotion data is used to evaluate abnormal behavior and optimize responses.

[1801] Optimizing responses based on perceived emotions

[1802] The server uses the emotion data obtained from the emotion engine to optimize the content of the alert and emergency response. For example, if a user is in a panic state, the server sends an alert that increases the level of urgency to the relevant authorities.

[1803] User Notifications

[1804] User devices (smartphones and PCs) receive notifications when abnormal behavior or emergencies occur, including a detailed description of the event, a link to a video clip, and appropriate countermeasures based on the user's feelings.

[1805] Cooperation with related organizations

[1806] The server waits for confirmation of the alert sent to the relevant authorities, and once confirmation is received, it provides the next important update (e.g., changes in the situation on the ground, additional video clips, etc.) in real time, allowing the relevant authorities to respond quickly and accurately.

[1807] Big Data Analysis and Prediction

[1808] The server uses the accumulated video data and emotion data for big data analysis, training predictive models to forecast crimes and emergencies, and supporting optimal resource allocation and countermeasure planning based on the prediction results.

[1809] Specific examples

[1810] Example 1: Incident detection on public transport

[1811] The device (surveillance camera) continues to monitor the video near the ticket gate of the station. Among the people passing by, it detects one person holding a knife.

[1812] The server detects this abnormal behavior and uses an emotion engine to analyze the emotions of the person wielding the knife and those around them. If emotions of panic or fear are recognized, a high-level alert is generated and sent to the relevant authorities.

[1813] The relevant authorities (police) receive a notification and dispatch the nearest police officer to the scene. The server continues to monitor the footage from the camera and provides additional information to the relevant authorities.

[1814] The user (station staff) will receive a notification on their smartphone, and will be able to immediately grasp the situation and prepare a response.

[1815] Example 2: Suspicious behavior detection in a commercial area

[1816] The device (surveillance camera) monitors the video at the entrance of the shopping mall and detects the behavior of a person stealing goods from multiple stores.

[1817] The server detects this abnormal behavior and uses an emotion engine to analyze the emotions of the thief and those around them. If emotions of anxiety or confusion are recognized, an alert is generated suggesting appropriate responses and sent to the relevant authorities.

[1818] The relevant authorities (police) receive the notification and immediately dispatch the nearest police officers to the scene. The server updates the relevant authorities as the situation changes.

[1819] The user (store staff) will receive a notification on their smartphone, immediately check the security camera footage, and take appropriate action while waiting for the police to arrive.

[1820] In this way, by combining an emotion engine with an AI-powered surveillance camera system, abnormal behavior can be detected in real time, strengthening collaboration with users and relevant agencies, and enabling swift and accurate responses. By utilizing big data analysis and predictive models, crime prevention and efficient resource allocation become possible.

[1821] The processing flow will be explained below.

[1822] Step 1:

[1823] The device (surveillance camera) captures real-time video of the specified area. The video is divided into frames and temporarily stored in a buffer.

[1824] Step 2:

[1825] The terminal (surveillance camera) compresses the video data in the buffer at regular intervals (e.g., every second) and sends it to a server via the Internet.

[1826] Step 3:

[1827] The server receives the video data sent from the device, decompresses it, and immediately inputs it into the AI ​​model.

[1828] Step 4:

[1829] The server analyzes the transmitted video data and identifies people and objects using an AI model that is optimized based on pre-trained data.

[1830] Step 5:

[1831] Based on the analysis results, the server tracks the movements of people and objects in the video and uses an abnormal behavior detection algorithm to detect abnormal behavior such as falls, fights, or the possession of knives.

[1832] Step 6:

[1833] When the server detects abnormal or emergency activity, it clips the frame of data and flags it as abnormal activity. The video clip, timestamp, and location information are saved.

[1834] Step 7:

[1835] The server uses an emotion engine to analyze the user's facial expressions and voice based on the analysis results of the video data and abnormal behavior, and recognizes emotions.

[1836] Step 8:

[1837] The server then optimizes the content of the alert and emergency response based on the recognized emotion. For example, if a user is in a panic state, it will generate a high-urgency alert and send it to the relevant authorities.

[1838] Step 9:

[1839] The server sends an alert to the relevant authorities (police, fire department) in the form of a push notification. The alert message includes details of the incident, a video clip, a timestamp, and location information.

[1840] Step 10:

[1841] The relevant authorities (police and fire departments) receive an alert notification and begin preparing to respond to the incident, dispatching the nearest police officers and firefighters to the scene.

[1842] Step 11:

[1843] The server monitors the response status of relevant agencies and continues to transmit video data in real time if the situation at the scene changes, providing the latest information.

[1844] Step 12:

[1845] The server sends notifications of emergencies or abnormal behavior to specific user devices (e.g., commercial facility staff, security personnel), including detailed descriptions, links to video clips, and appropriate countermeasures based on the recognized emotions.

[1846] Step 13:

[1847] Users (such as commercial facility staff) receive notifications on their smartphones or PCs, immediately check the situation, and take action to ensure the safety of their surroundings and prevent damage, if necessary.

[1848] Step 14:

[1849] The server stores the accumulated video data and emotion data and uses it for big data analysis, using predictive models to forecast future crimes and emergencies.

[1850] Step 15:

[1851] The server generates detailed reports based on the analysis results to help optimize resource allocation and plan crime prevention activities, and the generated reports enable efficient countermeasures to be taken.

[1852] This enables the system to utilize AI for real-time monitoring and immediate response, and employs an emotion engine to take appropriate action in the event of abnormal behavior or an emergency, significantly improving safety in cities, commercial districts, public transportation, and communities.

[1853] Example 2

[1854] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1855] Current surveillance camera systems are limited to simply monitoring video footage and detecting abnormal behavior, and lack the ability to accurately assess users' emotions and situations and respond quickly and appropriately. Furthermore, their functionality for coordinating with relevant agencies and preventing abnormal behavior is limited. This leads to problems such as delayed initial responses to emergencies and difficulty allocating appropriate resources.

[1856] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1857] In this invention, the server includes a means for acquiring camera footage, a means for analyzing the acquired camera footage in real time and identifying people or objects, a means for detecting abnormal behavior or emergencies, a means for automatically issuing an alert based on the detected abnormal behavior or emergency, a means for coordinating with relevant organizations in real time, and a means for recognizing emotions from the user's facial expressions and voice and optimizing responses. This integrates the process from detecting abnormal behavior to emergency response, enabling flexible responses based on the user's emotions. Furthermore, the combined use of big data analysis and predictive models enables more accurate prediction and prevention of abnormal behavior and emergencies, enabling optimal resource allocation and rapid response.

[1858] "Camera footage" refers to visual information obtained from a surveillance camera.

[1859] "Video data" refers to data that represents camera images in digital format.

[1860] "Real-time analysis" refers to processing and analyzing data as soon as it is received.

[1861] "Person or object identification" refers to analyzing surveillance camera footage and identifying people or objects present in the footage.

[1862] "Abnormal behavior" refers to behavior that deviates from general patterns of behavior and is considered dangerous or suspicious.

[1863] An "emergency" is a sudden situation or event that requires immediate action.

[1864] "Alert" refers to a warning or notification issued when abnormal behavior or an emergency is detected.

[1865] "Related agencies" refers to external organizations (e.g., police, fire department, etc.) that respond to abnormal behavior or emergencies.

[1866] "Facial expressions" refer to the movements and changes of facial muscles that express a person's emotions and psychological state.

[1867] "Audio" refers to data related to human voices and sounds.

[1868] "Emotion recognition" refers to identifying a user's emotions and mental state from their facial expressions and voice.

[1869] "Optimizing responses" refers to selecting and implementing the most appropriate and effective response measures based on emotional data and the situation.

[1870] "Big data analytics" refers to the process of analyzing large amounts of data and extracting meaningful information and patterns.

[1871] "Predictive model" refers to a mathematical or statistical model used to forecast future events or trends based on historical data.

[1872] The system of the present invention combines an emotion engine with a high-performance surveillance camera system that utilizes AI, and is capable of identifying people and objects, detecting abnormal behavior and emergency situations, automatically issuing alarms, and recognizing the user's emotions and responding based on those emotions. The following are modes for implementing the present invention.

[1873] System Configuration

[1874] The system includes the following main components:

[1875] 1. Terminal (surveillance camera)

[1876] 2. Server

[1877] 3. Emotion Engine

[1878] 4. User devices (smartphones and PCs)

[1879] 5. Related agencies (police, fire department, etc.)

[1880] Acquiring and transmitting camera images

[1881] The device (surveillance camera) captures real-time video of a specified area. For example, this applies to surveillance cameras installed in surveillance systems for public transportation or commercial districts. The video is divided into frames and temporarily stored in a buffer. The video data is periodically compressed using a compression technology such as H.264 and sent to a server using a secure protocol such as HTTPS.

[1882] Receiving and analyzing video data

[1883] The server receives the compressed video data sent from the device. The received data is decompressed and input into a generative AI model (e.g., YOLOv4 or OpenPose). This AI model identifies people and objects and detects abnormal behavior and emergencies. For example, it can detect dangerous behavior such as someone carrying a knife on public transportation.

[1884] Detecting abnormal behavior and issuing alerts

[1885] The server then assigns a timestamp and location information to any abnormal behavior or emergency detected by the AI ​​model. Based on this, an alert message is generated and sent to the relevant authorities in the form of a push notification, along with a detailed description of the incident and a link to a video clip. For example, police and fire departments are notified immediately.

[1886] Introducing an emotion engine and optimizing responses

[1887] The emotion engine analyzes the user's facial expressions and voice from the acquired video data and abnormal behavior analysis data to recognize their emotions. The server can use this emotion data to optimize the content of warnings and emergency responses. For example, if a user is in a panic state, it can send a high-level alert to the relevant authorities.

[1888] User notification and collaboration with relevant organizations

[1889] User devices (smartphones and PCs) receive notifications when abnormal behavior or emergencies occur. The notifications include a detailed description of the event, a link to a video clip, and appropriate countermeasures based on emotion data, allowing users to respond quickly.

[1890] Big Data Analysis and Prediction

[1891] The server then performs big data analysis on the accumulated video and emotion data, which can then be used to train predictive models to forecast future abnormal behavior and emergencies. The prediction results can then be used to optimize resource allocation and support countermeasure planning.

[1892] Resulting System Behavior

[1893] This system analyzes real-time footage from surveillance cameras, detects abnormal behavior and emergencies, and immediately coordinates with relevant agencies to take appropriate action. Furthermore, it has the ability to analyze user emotions and respond appropriately based on those emotions, enabling rapid and effective response to emergencies. Big data analysis also improves prediction accuracy, enabling advance countermeasures and optimal resource allocation.

[1894] Examples of prompt statements

[1895] "Please explain in detail the process of the system that detects abnormal behavior at station ticket gates and also analyzes the emotions of those around you."

[1896] "Please explain a system that detects suspicious behavior in a shopping mall and performs sentiment analysis based on that behavior."

[1897] In this way, the system of the present invention enables real-time detection of abnormal behavior, prompt notification to users and relevant organizations, and optimal response based on emotional data.

[1898] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1899] Step 1:

[1900] The terminal (surveillance camera) acquires real-time video of the designated area.

[1901] Input: Video of the specified area

[1902] Processing: Capture video in frames (e.g. 30 frames per second) and save as frame data

[1903] Output: Frame data

[1904] Step 2:

[1905] The terminal (surveillance camera) stores the acquired frame data in a buffer, compresses it at regular intervals, and sends it to the server.

[1906] Input: Frame data

[1907] Processing: Frame data is temporarily stored in a buffer and compressed using a compression algorithm such as H.264

[1908] Output: Compressed video data is sent to the server

[1909] Step 3:

[1910] The server receives the compressed video data sent from the terminal and decompresses it.

[1911] Input: Compressed video data

[1912] Processing: Decompresses received data and restores it as frame data

[1913] Output: Decompressed frame data

[1914] Step 4:

[1915] The server inputs the decompressed frame data into a generative AI model to identify people or objects.

[1916] Input: Decompressed frame data

[1917] Processing: Input frame data into a generative AI model (e.g., YOLOv4, OpenPose, etc.) to identify people and objects.

[1918] Output: Data of identified people or objects (location, label, etc.)

[1919] Step 5:

[1920] The server analyzes the data of the identified person or object to detect abnormal behavior or emergency situations.

[1921] Input: Identified person or object data

[1922] Processing: Uses analytical algorithms to detect abnormal behavior and emergencies

[1923] Output: Data on detected abnormal behavior or emergency (e.g., timestamp, location, behavior type, etc.)

[1924] Step 6:

[1925] The server issues alerts based on detected abnormal behavior or emergency situations and sends push notifications to relevant authorities.

[1926] Input: Data on abnormal behavior and emergencies

[1927] Processing: Generate an alert message and send it to the relevant authorities in the form of a push notification.

[1928] Output: Alert message to relevant organizations

[1929] Step 7:

[1930] The emotion engine uses the acquired video data and abnormal behavior analysis data to recognize emotions from the user's facial expressions and voice.

[1931] Input: Video data, abnormal behavior analysis data

[1932] Processing: Apply emotion recognition algorithms to identify the user's emotions

[1933] Output: Recognized emotion data (e.g., panic, fear, anxiety, etc.)

[1934] Step 8:

[1935] The server uses the recognized emotion data to optimize alert content and emergency response.

[1936] Input: Emotion data

[1937] Processing: Adjust the content and urgency of the warning message and optimize the necessary response measures.

[1938] Output: Optimized warning messages and countermeasures

[1939] Step 9:

[1940] User devices (smartphones and PCs) receive notifications when abnormal behavior or emergencies occur.

[1941] Input: Optimized warning message

[1942] Action: Display a notification and inform the user of the situation

[1943] Output: A message to inform the user

[1944] Step 10:

[1945] The server waits for confirmation of receipt of alerts sent to relevant authorities and provides important updates in real time.

[1946] Input: Acknowledgments from relevant agencies and new video data

[1947] Processing: Get acknowledgements and real-time updates on important information like changing conditions on the ground and additional footage clips

[1948] Output: Update information for relevant organizations

[1949] Step 11:

[1950] The server uses the accumulated video data and emotion data for big data analysis.

[1951] Input: Accumulated video data, emotion data

[1952] Processing: Data is stored in a data warehouse and analyzed using data mining techniques

[1953] Output: Analysis results and training data for predictive models

[1954] Step 12:

[1955] The server uses predictive models to forecast crimes and emergencies and help plan countermeasures.

[1956] Input: Prediction model, analysis results

[1957] Processing: Applying predictive algorithms to predict future abnormal behavior and emergencies

[1958] Output: Prediction results and supporting information for action plans

[1959] (Application example 2)

[1960] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1961] Conventional surveillance systems have the ability to identify people and objects and detect abnormal behavior and emergencies, but they are unable to recognize user emotions and provide responses based on those emotions. Furthermore, responses after detecting abnormal behavior are limited, often resulting in failure to contact relevant authorities or propose appropriate countermeasures. To address these issues, the present invention aims to provide a system that adds emotion recognition functionality and optimizes responses after detecting abnormal behavior.

[1962] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for acquiring camera images for identifying people or objects, means for analyzing camera image data and identifying people or objects, means for detecting abnormal behavior or emergency situations, and means for recognizing user emotions and optimizing responses based on the emotions. This makes it possible to take appropriate action based not only on identification but also on emotion data when abnormal behavior or an emergency occurs.

[1963] "Person or object identification" refers to technology that analyzes camera video data to identify people or objects.

[1964] "Unusual behavior or emergency" means any behavior that deviates from normal patterns of behavior or any situation that poses an immediate danger to life or property.

[1965] "Automatic alerting means" means a technology that automatically issues an alert in accordance with pre-defined conditions based on detected abnormal behavior or emergency situations.

[1966] "Means for recognizing the user's emotions and optimizing the response based on that" refers to technology that analyzes the user's emotions using acquired video and audio data and provides the optimal response based on the results.

[1967] "Means of coordinating with relevant agencies in real time" refers to technology that allows for instant communication with relevant agencies such as the police and fire department, providing necessary information, and maintaining a cooperative relationship.

[1968] "Big data analytics and predictive models" are techniques for analyzing large amounts of data to derive specific patterns and trends, which can be used to predict abnormal behavior and emergencies.

[1969] "Means for providing video clips, timestamps, and location information" refers to technology that, when issuing an alert, adds detailed information such as specific video data, the time of occurrence, and location.

[1970] "Means for proposing appropriate countermeasures" refers to technology that presents optimal and effective countermeasures based on detected abnormal behavior and user emotional data.

[1971] System Overview

[1972] The present invention is a system that identifies people and objects, detects abnormal behavior and emergencies, and automatically issues an alarm. Furthermore, it recognizes the user's emotions and provides an optimal response based on those emotions, making the response after abnormal behavior is detected more effective.

[1973] System Components

[1974] The system includes the following major components:

[1975] 1. Surveillance cameras

[1976] 2. Server

[1977] 3. Emotion Recognition Engine

[1978] 4. User devices (smartphones and PCs)

[1979] 5. Related agencies (police, fire department, etc.)

[1980] System Operation

[1981] Acquiring and transmitting camera images

[1982] The surveillance camera captures real-time video of a specified area. The video is divided into frames and temporarily stored in a buffer. The video data in the buffer is compressed at regular intervals and sent to the server.

[1983] Receiving and analyzing video data

[1984] The server receives video data transmitted from the surveillance cameras. The received data is decompressed and immediately input into a generative AI model. This model identifies people and objects, and detects abnormal behavior and emergencies. The detected data is then recorded in a database.

[1985] Detecting abnormal behavior and issuing alerts

[1986] The server captures and flags any abnormal behavior or emergency situations, capturing their timestamp and location information. The video clip, timestamp, and location information are saved, and an alert message is generated and sent to the appropriate authorities in the form of a push notification.

[1987] Using the Emotion Recognition Engine

[1988] The server uses the acquired video data and abnormal behavior analysis data to recognize emotions from the user's facial expressions and voice using an emotion recognition engine. This emotion data is used to evaluate abnormal behavior and optimize responses.

[1989] Optimizing responses based on perceived emotions

[1990] The server uses the emotion data obtained from the emotion recognition engine to optimize the content of the alert and emergency response, for example, sending an alert to the relevant authorities that increases the level of urgency if the user is in a panic.

[1991] User notification and response

[1992] The user device receives a notification when abnormal behavior or an emergency occurs. The notification includes a detailed description of the event, a link to a video clip, and appropriate countermeasures based on the user's emotions. The notification arrives on the user's smartphone, allowing the user to immediately understand the situation and prepare a response.

[1993] Hardware and software used

[1994] Hardware:

[1995] Surveillance cameras (IP cameras, etc.)

[1996] Smartphone or PC

[1997] software:

[1998] OpenCV: Camera image acquisition and analysis

[1999] Keras: Loading and predicting emotion recognition models

[2000] requests: Server alert notification

[2001] Specific processing examples

[2002] Example 1: Suspicious behavior detection in a shopping mall

[2003] 1. A surveillance camera monitors footage at the entrance to a shopping mall and detects an individual stealing items from multiple stores.

[2004] 2. The server detects this abnormal behavior and uses an emotion recognition engine to analyze the emotions of the thief and those around them. If emotions of anxiety or confusion are recognized, an alert is generated suggesting appropriate responses and sent to the relevant authorities.

[2005] 3. The relevant authorities (police) receive the notification and immediately dispatch the nearest police officers to the scene. The server updates the relevant authorities with information as the situation changes.

[2006] 4. The user (store staff) receives a notification on their smartphone, immediately checks the security camera footage, and takes appropriate action while waiting for the police to arrive.

[2007] Prompt Sentence Examples

[2008] "Detect anomalous behavior in this camera footage and recognize the emotions of people around it. If anomalous behavior is detected, generate an alert message suggesting appropriate countermeasures based on the emotional data."

[2009] The above is an embodiment of the present invention. The present invention makes it possible to detect abnormal behavior in real time, strengthen cooperation with users and related organizations, and realize a prompt and accurate response.

[2010] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[2011] Step 1:

[2012] The surveillance camera captures real-time video of a specified area. The input is the video data from the surveillance camera, and the output is the video frame stored in the buffer. This frame is temporarily stored for further processing.

[2013] Step 2:

[2014] The server compresses the video data in the buffer at regular intervals and sends it to the server. The input is the video frames stored in the buffer, and the output is the compressed video data. Data compression enables efficient use of network bandwidth.

[2015] Step 3:

[2016] The server receives the compressed video data and decompresses it. The input is the compressed video data, and the output is the decompressed video data. The decompressed data is immediately input into the generative AI model.

[2017] Step 4:

[2018] The server uses a generative AI model to identify people and objects. The input is the decompressed video data, and the output is the information of the identified people and objects. The generative AI model identifies people and objects based on video analysis.

[2019] Step 5:

[2020] The server further analyzes the information of the identified people and objects to detect abnormal behavior or emergency situations. The input is the information of the identified people and objects, and the output is a flag indicating abnormal behavior or emergency situation. An algorithm is used to detect patterns of abnormal behavior or emergency situations.

[2021] Step 6:

[2022] When an abnormal behavior or emergency is detected, the server acquires the timestamp and location information and sets a flag. The input is the flag for abnormal behavior detection or emergency detection, and the output is the timestamp and location information.

[2023] Step 7:

[2024] The server generates an alert message based on the video clip, timestamp, and location information, and sends it to the relevant organizations. The input is the timestamp, location information, and video clip, and the output is the generated alert message. It is sent to the relevant organizations in the form of a push notification.

[2025] Step 8:

[2026] The server uses an emotion recognition engine to analyze the user's facial expressions and voice based on the acquired video data and recognizes their emotions. The input is video data and the output is emotion data. The emotion recognition engine estimates the user's emotions from their facial expressions and voice.

[2027] Step 9:

[2028] The server optimizes the content of the alert and emergency response using the emotion data obtained from the emotion recognition engine. The input is emotion data, and the output is optimized alert content and emergency response measures. The urgency and content of the response are adjusted based on the emotion data.

[2029] Step 10:

[2030] The user device receives notifications when abnormal behavior or an emergency occurs. The input is the optimized alert content and emergency response measures, and the output is a notification displayed on the user device. The user receives the notification on their smartphone, allowing them to immediately understand the situation and prepare a response.

[2031] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[2032] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[2033] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[2034] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[2035] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[2036] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[2037] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[2038] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, motorcycles, and other devices, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[2039] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[2040] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[2041] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[2042] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[2043] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory s...

Claims

1. means for acquiring camera footage for identifying people or objects; A means for analyzing camera image data and identifying a person or object; means for detecting abnormal behavior or emergency situations; means for automatically issuing an alert based on the detected abnormal behavior or emergency situation; A system that includes a means of coordinating with relevant agencies in real time.

2. The system of claim 1 , further comprising means for predicting and preventing abnormal behavior or emergency situations using big data analysis and predictive models.

3. 10. The system of claim 1, further comprising means for providing detailed information including a video clip, a timestamp, and location information when an alert is sent to appropriate authorities.

4. 10. The system of claim 1, further comprising means for sending a notification of abnormal activity or an emergency to a user terminal, providing a detailed description and a link to a video clip.

5. 10. The system of claim 1, further comprising means for compressing and transmitting camera video data.

Images