system

Abstract

Provide a system. An image acquisition means for collecting camera data for detecting anomalies; an analysis means for detecting anomalies from the acquired camera data; a notification means for notifying an abnormality detected by the analysis means; A system including:

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] In public places and educational institutions, it is important to detect suspicious behavior and the bringing in of dangerous items early and respond quickly. However, conventional monitoring systems have difficulty in detecting abnormalities in real time, and since notifications after detecting an abnormality are generally handled manually, there is a risk of delayed response. This increases the likelihood of situations threatening human life and safety. Furthermore, educational institutions lack sufficient systems to prevent children and students from accidentally leaving the premises, which could also lead to serious accidents. [Means for solving the problem]

[0005] The present invention is a system that includes a video acquisition means for collecting camera data for detecting abnormalities, an analysis means for detecting abnormalities from the acquired camera data, and a notification means for notifying abnormalities detected by the analysis means. Specifically, a camera or sensor is used as the video acquisition means to collect data in real time, and the analysis means utilizes object recognition algorithms and behavior recognition algorithms to detect knives, suspicious liquids, and suspicious behavior. If an abnormality is detected through this analysis, the notification means notifies staff and police in real time. This automates a process that was previously performed manually, enabling a rapid and accurate response. This system also includes a function to prevent children and students from unintentionally going outside in educational institutions, thereby improving safety.

[0006] "Video acquisition means" refers to a device that collects video data in real time using a camera, sensor, etc.

[0007] "Analysis means" refers to algorithms or software for detecting abnormalities based on acquired video data, including, for example, object recognition algorithms and behavior recognition algorithms.

[0008] "Notification methods" are systems for transmitting information in real time to relevant parties such as staff and police when an abnormality is detected, and include email, SMS, in-app notifications, etc.

[0009] An "object recognition algorithm" is a program that uses machine learning and image processing technology to detect specific objects, such as knives or suspicious liquids, from video data acquired by a camera.

[0010] A "behavior recognition algorithm" is a program that uses machine learning and pattern recognition technology to analyze human movements and actions from video data and detect suspicious behavior.

[0011] "Personnel" refers to the person responsible for responding to any abnormalities detected at the location where the system is installed.

[0012] "Real-time" refers to extremely short delays between data collection and response, meaning that information is processed and notified almost instantly.

[0013] "Abnormal" refers to behavior or conditions that deviate from normal circumstances, including possession of a knife, possession of suspicious liquids, suspicious behavior, or leaving the facility without permission. [Brief explanation of the drawings]

[0014] [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

[0015] 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.

[0016] First, the terms used in the following description will be explained.

[0017] 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).

[0018] 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.

[0019] 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.

[0020] 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.

[0021] 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."

[0022] [First embodiment]

[0023] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0024] 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.

[0025] 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).

[0026] 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.

[0027] 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.

[0028] 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.

[0029] 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.

[0030] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0031] 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.

[0032] 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.

[0033] 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.

[0034] 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."

[0035] This invention is a system that quickly detects abnormalities in public places and educational institutions and automates appropriate responses. This system combines video capture, analysis, and notification means to significantly improve safety at the site.

[0036] Program processing

[0037] 1. Data Collection

[0038] Terminal: Collects real-time video from surveillance cameras and sensors installed in various locations. For example, a camera installed on a station platform captures video of the entire platform.

[0039] Terminal: Temporarily stores collected video data and makes it available for immediate analysis. For example, video captured by a camera is temporarily stored in local storage.

[0040] 2. Data Preprocessing

[0041] Server: Filters the received video data to remove noise, converting it into a clear, easy-to-analyze image.

[0042] Server: Converts the filtered video data to the optimal resolution and prepares it for sending to the analysis means.

[0043] 3. Analysis for Anomaly Detection

[0044] Server: The filtered video data is fed into an AI model, where object and behavior recognition algorithms are applied, for example to detect people holding knives or handling suspicious liquids.

[0045] Server: Identifies specific abnormal behavior, such as detecting someone walking quickly or moving differently from the normal crowd.

[0046] 4. Identifying abnormal situations

[0047] Server: If an anomaly is detected based on the analysis results, a detailed report is generated, including the location, time, and type of anomaly detected.

[0048] Server: Transforms the report into a readable format and prepares it for notification.

[0049] 5. Notification Processing

[0050] Server: Based on the generated report, it immediately notifies officials and police via email, SMS, in-app alerts, etc.

[0051] Notification method: A notification is sent to the employee's device and an alert sound is generated to attract attention. For example, a notification is displayed on the computer of an employee working in the monitoring room.

[0052] 6. On-site response

[0053] User: Upon receiving the notification, staff or police rush to the scene and check the details of the abnormality. For example, staff rush to the scene while checking the surveillance camera footage.

[0054] User: Eliminate the cause of the anomaly and take appropriate action to return the situation to safety.

[0055] 7. Video storage and post-processing

[0056] Server: Stores video footage of detected anomalies for later analysis or as evidence. For example, it stores video footage of before and after the anomaly occurred in a specific format.

[0057] Server: Analyzes the stored video data and uses it to evaluate and improve system performance.

[0058] Specific examples

[0059] Example on a station platform:

[0060] Terminal: Cameras installed on station platforms capture passenger activity in real time. For example, cameras can capture the situation of large numbers of passengers on the platform during rush hour.

[0061] Server: Receives video data and analyzes abnormal behavior, such as detecting someone pulling out a knife in a crowd.

[0062] Notification method: If an abnormality is detected, an emergency notification will be sent to station staff and the police via email or SMS. For example, a staff member may receive a notification on their smartphone stating that "a person with a knife is in the center of the platform."

[0063] User: Staff rush to the scene and contact the police. For example, staff immediately head to the scene and handle the situation until the police arrive.

[0064] Thus, the present invention is a system that provides an embodiment for quickly detecting abnormalities in public places and educational institutions and taking appropriate measures.

[0065] The processing flow will be explained below.

[0066] Step 1:

[0067] Terminal: Collects real-time video from surveillance cameras and sensors installed in various locations. For example, a camera installed on a station platform captures video of the entire platform.

[0068] Step 2:

[0069] Terminal: Collected video data is temporarily stored in local storage, making it available for immediate analysis. For example, video captured by a camera is temporarily stored in a buffer.

[0070] Step 3:

[0071] Server: Performs filtering to remove noise from the received video data, for example, removing non-essential parts of the video data and converting it into a clear, easy-to-analyze format.

[0072] Step 4:

[0073] Server: Converts filtered video data to the optimal resolution and prepares it for sending to the analysis algorithm. Converts data into the format required by the analysis method.

[0074] Step 5:

[0075] Server: The filtered video data is fed into an AI model, where object and behavior recognition algorithms are applied, for example to detect people holding knives or handling suspicious liquids.

[0076] Step 6:

[0077] Server: Identifies specific abnormal behavior, such as people walking quickly or moving differently from the general crowd, thereby identifying suspicious individuals in real time.

[0078] Step 7:

[0079] Server: If an anomaly is detected based on the analysis results, a detailed report is generated, including the location, time, and type of anomaly detected.

[0080] Step 8:

[0081] Server: Transforms reports into a format that is easy to read and prepares them for notification, e.g., creating text and visual reports.

[0082] Step 9:

[0083] Server: Based on the generated report, it immediately notifies officials and police via email, SMS, in-app alerts, etc.

[0084] Step 10:

[0085] Terminal: A notification is sent to the employee's terminal and an alert sound is generated to attract attention. For example, a notification is displayed on the computer of an employee working in the monitoring room.

[0086] Step 11:

[0087] User: Upon receiving the notification, staff or police rush to the scene and check the details of the abnormality. For example, staff rush to the scene while checking the surveillance camera footage.

[0088] Step 12:

[0089] User: Eliminate the cause of the anomaly and take appropriate action to return the situation to safety. For example, staff will seize the dangerous material and coordinate with the police.

[0090] Step 13:

[0091] Server: Stores video footage of detected anomalies for later analysis or as evidence. For example, it stores video footage of before and after the anomaly occurred in a specific format.

[0092] Step 14:

[0093] Server: Analyzes the stored video data to evaluate and improve system performance, for example by retraining the AI ​​model based on past cases.

[0094] Example 1

[0095] 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."

[0096] Public places and educational institutions require the ability to quickly and accurately detect anomalies and automate appropriate responses. However, conventional systems lack the accuracy of anomaly detection and notification speed required, making manual monitoring and response necessary. Furthermore, the process of storing detected anomalies and post-analysis was often inefficient. A system that solves these problems and significantly improves safety is needed.

[0097] 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.

[0098] In this invention, the server includes a data acquisition means for collecting video data for detecting anomalies, a data preprocessing means for removing noise from the acquired video data and optimizing resolution, an analysis means for detecting anomalies from the video data processed by the data preprocessing means, a notification means for notifying anomalies detected by the analysis means, and a video storage means for storing video of the detected anomalies for use as later analysis or evidence. This enables rapid and accurate detection of anomalies, real-time notification and storage, and further subsequent analysis.

[0099] "Data acquisition means" refers to the devices and technologies that collect the video data necessary to detect abnormalities.

[0100] "Data pre-processing means" refers to technology or devices that perform processing to remove noise from acquired video data and optimize resolution.

[0101] "Analysis means" refers to a device that includes algorithms and techniques for detecting anomalies based on pre-processed video data.

[0102] "Notification means" refers to the technology or device used to notify relevant parties of abnormalities detected by the analysis means.

[0103] "Video storage means" refers to technology or equipment for storing video of detected abnormalities and using it for later analysis or as evidence.

[0104] "Object recognition algorithm" refers to the computational processes and techniques used to detect and identify objects based on video data.

[0105] "Behavior recognition algorithm" refers to the computational processes and technologies used to detect and identify behaviors based on video data and determine abnormal behavior.

[0106] "Emergency notification function" refers to technology and systems that notify relevant parties such as staff and police in real time of details of detected abnormalities.

[0107] The present invention relates to a system for quickly detecting abnormalities in public places and educational institutions and automating appropriate responses. The system includes a data acquisition unit, a data preprocessing unit, an analysis unit, a notification unit, and a video storage unit.

[0108] The system is programmed as follows:

[0109] First, surveillance cameras and other data acquisition methods installed at various locations collect video data in real time. For example, cameras installed on train platforms capture passenger movements.

[0110] The device then temporarily stores the collected video data and makes it available for immediate analysis using local storage.

[0111] The server then performs a filtering process to remove noise from the received video data, making the video clearer and easier to analyze. The filtered video data is then converted to an optimal resolution and prepared for transmission to the analysis means.

[0112] The analysis method inputs the filtered video data into an AI model and applies object and behavior recognition algorithms to detect, for example, people holding knives or handling suspicious liquids, as well as identify specific abnormal behaviors, such as people walking quickly or moving differently from the general crowd.

[0113] If the analysis results in an anomaly, the server generates a detailed report that includes the location, time, and type of anomaly detected, and converts the report into an easy-to-read format for notification.

[0114] The notification method will immediately notify employees and police based on the contents of the generated report. Notifications will be sent via email, SMS, in-app alerts, etc. For example, an emergency notification such as "There is a person with a knife on the station platform" will be sent to an employee's smartphone.

[0115] The users who receive the notification, i.e., officials or police, rush to the scene and check the details of the abnormality. Officials rush to the scene while checking the surveillance camera footage and take appropriate action to eliminate the cause of the abnormality.

[0116] Finally, the server stores the video of the detected anomaly for later analysis and as evidence. The stored video data can be analyzed to evaluate and improve the system's performance.

[0117] Prompt Sentence Examples

[0118] Below are some example prompts to input to the generative AI model:

[0119] "Please explain the most effective method for detecting abnormal behavior in public places."

[0120] "What are the key elements of a system that quickly detects anomalies and automatically notifies you?"

[0121] Thus, the present invention is a system that provides an embodiment for quickly detecting abnormalities in public places and educational institutions and taking appropriate measures.

[0122] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0123] Step 1: Data collection

[0124] Terminal: Surveillance cameras installed in various locations collect video data in real time. For example, cameras installed on train platforms capture passenger movements.

[0125] Input: Real-time video data from public places and educational institutions.

[0126] Output: Real-time video data.

[0127] Step 2: Temporarily save the video data

[0128] Terminal: Collected video data is temporarily stored in local storage. For example, video data captured by a camera is temporarily stored on a local drive so that it can be analyzed immediately.

[0129] Input: Real-time video data.

[0130] Output: Stored temporary video data.

[0131] Step 3: Filtering

[0132] Server: Filters the received video data to remove noise. By removing unnecessary noise and blur, the video becomes clear and easier to analyze.

[0133] Input: Stored temporary video data.

[0134] Output: Filtered video data.

[0135] Step 4: Optimize the resolution

[0136] Server: Converts the filtered video data to the appropriate resolution and prepares it for analysis. For example, compresses high-resolution video slightly to maintain analysis accuracy and speed.

[0137] Input: Filtered video data.

[0138] Output: Optimized resolution video data.

[0139] Step 5: Input to the AI ​​model

[0140] Server: The optimized resolution video data is input into the AI ​​model and an object recognition algorithm is applied to detect people and objects in the video data.

[0141] Input: Video data with optimized resolution.

[0142] Output: Object and person detection results.

[0143] Step 6: Applying the activity recognition algorithm

[0144] Server: Uses the results of the object recognition algorithm to apply the behavior recognition algorithm, for example, to detect people whose movements deviate from their normal behavioral patterns.

[0145] Input: Object and person detection results.

[0146] Output: Abnormal behavior detection results.

[0147] Step 7: Generate anomaly detection reports

[0148] Server: If the analysis results in an anomaly, a detailed report is generated, including the location, time, and specific details of the anomaly.

[0149] Input: Anomalous behavior detection results.

[0150] Output: Anomaly detection report.

[0151] Step 8: Convert the report to another format

[0152] Server: Converts the generated reports into a format that is easy to read, for example, creating reports in PDF or HTML format.

[0153] Input: Anomaly detection report.

[0154] Output: Formatted report.

[0155] Step 9: Call Emergency Services

[0156] Server: Based on the generated reports, it provides instant notifications to officials and police via email, SMS or in-app alerts.

[0157] Input: The formatted report.

[0158] Output: Emergency call notification.

[0159] Step 10: Displaying notifications

[0160] Notification method: An emergency notification will be displayed on the employee's terminal and an alert sound will be emitted to attract attention. For example, the notification will be displayed immediately on the computer of an employee on duty in the monitoring room.

[0161] Input: Emergency call notification.

[0162] Output: Notifications that appear on the device.

[0163] Step 11: Rush to the scene

[0164] User: Upon receiving the notification, staff and police rush to the scene to check the details of the abnormality. Staff head to the scene while checking the surveillance camera footage.

[0165] Input: Notifications that appear on your device.

[0166] Output: Check the details of the error.

[0167] Step 12: Dealing with abnormalities

[0168] User: Eliminate the cause of the anomaly and take appropriate action to return the situation to safety. For example, staff may apprehend the anomaly.

[0169] Input: Check the details of the abnormality.

[0170] Output: Safe state restored.

[0171] Step 13: Save the abnormal video

[0172] Server: Stores video footage of detected anomalies for later analysis and as evidence. Stores footage of several minutes before and after the anomaly occurs.

[0173] Input: Detailed confirmation of anomaly and video data.

[0174] Output: Archived footage of the anomaly.

[0175] Step 14: Performance evaluation and system improvement

[0176] Server: Analyzes the stored video data to evaluate and improve the system's performance. For example, it evaluates detection accuracy and notification speed, and improves the AI ​​model.

[0177] Input: Archived anomalous footage.

[0178] Output: System improvement measures and evaluation results.

[0179] (Application example 1)

[0180] 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."

[0181] In modern society, there is a demand for immediate detection and rapid response of abnormal behavior in public places and educational institutions. However, current systems can be slow to detect and notify abnormalities, or on-site staff may not be able to respond immediately. Furthermore, there is a lack of means to immediately communicate detailed information about abnormalities to staff visually and audibly, making it difficult to respond to abnormalities efficiently. The objective of the present invention is to solve these problems.

[0182] 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.

[0183] In this invention, the server includes a video acquisition means for collecting camera data for detecting abnormalities, an analysis means for detecting abnormalities from the acquired camera data, a notification means for notifying the user of the abnormality detected by the analysis means, and a wearable display means for immediately conveying the notification information from the notification means to the user's visual and auditory senses. This makes it possible to quickly detect abnormal behavior, notify staff in real time, and respond immediately on site.

[0184] "Video acquisition means" refers to equipment and software that collects camera data for detecting abnormalities.

[0185] "Analysis means" refers to algorithms and software for detecting anomalies in acquired camera data.

[0186] The "notification means" refers to a system and software for notifying staff and other relevant parties of abnormalities detected by the analysis means.

[0187] The "wearable display means" is a device for instantly conveying the notification information of the notification means to the user's visual and auditory senses.

[0188] An "object recognition algorithm" is a program for identifying specific objects (e.g., knives, suspicious liquids) from video data.

[0189] A "behavior recognition algorithm" is a program that analyzes people's movements from video data and detects abnormal behavior (e.g., suspicious movements).

[0190] The "emergency notification function" is a function that notifies staff and police in real time of detected abnormalities.

[0191] A "server" is a main computing device that performs a series of processes such as data collection, analysis, and notification.

[0192] "User" refers to the staff and other relevant parties who receive and respond to the notified abnormality information.

[0193] The system for realizing this invention is configured using the following hardware and software: The hardware used is a surveillance camera, a cloud server, and smart glasses, while the software used is an AI model for video analysis (e.g., YOLO, OpenPose), a library for video noise reduction and filtering (e.g., OpenCV), and a notification service API (e.g., Firebase Cloud Messaging, Twilio).

[0194] The server collects video data from surveillance cameras in real time, which is first pre-processed with a program to remove noise and perform filtering, improving the quality of the video data and making it easier to analyze.

[0195] The server then inputs the preprocessed video data into an AI model and applies object and behavior recognition algorithms, such as YOLO to detect knives and suspicious liquids and OpenPose to identify abnormal behaviors, enabling the detection of specific abnormal behaviors and objects in real time.

[0196] If the server detects an anomaly based on these analysis results, it generates a detailed report containing information such as the location, time, and type of anomaly. After the report is generated, an emergency notification is sent immediately to staff and relevant parties via the notification service API. Notifications are sent in multiple ways, including email, SMS, and in-app alerts.

[0197] The user, a staff member, receives notifications through the smart glasses. The wearable display immediately conveys the notification information visually and audibly, allowing the staff member to quickly grasp the situation on-site. The staff member can check the details of any abnormalities in real time and take prompt action.

[0198] Examples:

[0199] For example, if a surveillance camera installed on a station platform detects a person holding a knife in the crowd, the server will filter the footage and identify the person as holding a knife using the YOLO model. If an anomaly is detected, the server will generate a detailed report and send a notification to the employee's smart glasses via the notification service API, stating, "There is a person holding a knife in the center of the platform." The employee will receive the notification and confirm the visual and audio alert, allowing them to respond to the scene quickly.

[0200] Example prompt sentence:

[0201] "Analyze the footage from the center of the platform to detect if there is anyone holding a knife."

[0202] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0203] Step 1:

[0204] The device transmits real-time video data captured by the camera to the cloud server. The input is video data from the surveillance camera, and the output is temporarily stored video data. This process involves acquiring video data and immediately transferring it to the server.

[0205] Step 2:

[0206] The server performs filtering to remove noise from the received video data. The input is the temporarily stored video data, and the output is the noise-removed video data. This process uses the OpenCV library to process the data to improve the video quality.

[0207] Step 3:

[0208] The server converts the filtered video data to the optimal resolution and prepares it for transmission to the analysis means. The input is the noise-removed video data, and the output is video data converted to a resolution suitable for analysis. This process involves data calculations using high-performance image processing algorithms.

[0209] Step 4:

[0210] The server then inputs the converted video data into an AI model and applies object and action recognition algorithms. The input is video data suitable for analysis, and the output is detected anomaly information. This process uses YOLO and OpenPose for object and action recognition and data calculation.

[0211] Step 5:

[0212] The server determines anomalies based on the analysis results of the AI ​​model and generates a detailed report. The input is the detected anomaly information, and the output is a report summarizing the details of the anomaly. This process generates a document summarizing the type of anomaly, the location where it occurred, the time, etc.

[0213] Step 6:

[0214] Based on the generated report, the server sends emergency notifications to staff and related parties via the notification service API. The input is a detailed report, and the output is the sent notification. This process uses services such as Firebase Cloud Messaging and Twilio to send notifications immediately.

[0215] Step 7:

[0216] The user receives notifications through smart glasses and acknowledges visual and audible alerts. The input is the notification sent, and the output is the user's acknowledgement and immediate response. This process uses the wearable device to display information and provide alerts.

[0217] Step 8:

[0218] After checking the contents of the notification, the user quickly heads to the site, checks the details of the abnormality, and takes action. The input is the content of the notification and the situation at the site, and the output is the appropriate action to eliminate the cause of the abnormality. This process allows for a quick response and problem resolution at the site.

[0219] 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.

[0220] This invention is a system that quickly detects abnormalities in public places and educational institutions and automates appropriate responses. This system combines video acquisition means, analysis means, notification means, and an emotion engine that recognizes user emotions, significantly improving safety.

[0221] Program processing

[0222] 1. Data Collection

[0223] Terminal: Collects real-time video from surveillance cameras and sensors installed in various locations. For example, a camera installed on a station platform captures video of the entire platform.

[0224] Terminal: Collected video data is temporarily stored in local storage, making it available for immediate analysis. For example, video captured by a camera is temporarily stored in a buffer.

[0225] 2. Data Preprocessing

[0226] Server: Performs filtering to remove noise from the received video data, for example, removing non-essential parts of the video data and converting it into a clear, easy-to-analyze format.

[0227] Server: Converts filtered video data to the optimal resolution and prepares it for sending to the analysis algorithm. Converts data into the format required by the analysis method.

[0228] 3. Analysis for Anomaly Detection

[0229] Server: The filtered video data is fed into an AI model, where object and behavior recognition algorithms are applied, for example to detect people holding knives or handling suspicious liquids.

[0230] Server: Identifies specific abnormal behavior, such as people walking quickly or moving differently from the general crowd, thereby identifying suspicious individuals in real time.

[0231] 4. Analysis for Emotion Recognition

[0232] Server: Runs an emotion engine based on video data and analyzes the user's facial expressions, voice, posture, etc. For example, it analyzes emotions such as smile, anger, and sadness from camera footage.

[0233] Server: Emotion data determined by the emotion engine is supplied to the analysis means and used as data to improve the accuracy of anomaly detection. For example, if a person with an angry expression is holding a knife, the degree of anomaly is set high.

[0234] 5. Identifying abnormal situations

[0235] Server: If an anomaly is detected based on the analysis results, a detailed report is generated, including the location, time, and type of anomaly detected.

[0236] Server: Transforms reports into a format that is easy to read and prepares them for notification, e.g., creating text and visual reports.

[0237] 6. Notification Processing

[0238] Server: Based on the generated report, it immediately notifies officials and police via email, SMS, in-app alerts, etc.

[0239] Notification method: A notification is sent to the employee's device and an alert sound is generated to attract attention. For example, a notification is displayed on the computer of an employee working in the monitoring room.

[0240] 7. On-site response

[0241] User: Upon receiving the notification, staff or police rush to the scene and check the details of the abnormality. For example, staff rush to the scene while checking the surveillance camera footage.

[0242] User: Eliminate the cause of the anomaly and take appropriate action to return the situation to safety. For example, staff will seize the dangerous material and coordinate with the police.

[0243] 8. Video storage and post-processing

[0244] Server: Stores video footage of detected anomalies for later analysis or as evidence. For example, it stores video footage of before and after the anomaly occurred in a specific format.

[0245] Server: Analyzes the stored video data to evaluate and improve system performance, for example by retraining the AI ​​model based on past cases.

[0246] Specific examples

[0247] Example on a station platform:

[0248] Terminal: Cameras installed on station platforms capture passenger activity in real time. For example, cameras can capture the situation of large numbers of passengers on the platform during rush hour.

[0249] Server: After receiving the video data, the server first uses object recognition and behavior recognition algorithms to detect people holding a knife or making suspicious movements.Then, the server analyzes the person's emotions based on facial expressions captured by the emotion engine.

[0250] Server: A person holding a knife and with an angry expression is judged to be abnormal.

[0251] Notification method: If an abnormality is detected, an emergency notification will be sent to station staff and the police via email or SMS. For example, a staff member may receive a notification on their smartphone saying, "There is an angry person holding a knife in the middle of the platform."

[0252] User: Staff rush to the scene and contact the police. For example, staff immediately head to the scene and handle the situation until the police arrive.

[0253] In this way, by combining emotion engines, the present invention is a system that provides an embodiment for detecting anomalies with higher accuracy and taking prompt and appropriate action.

[0254] The processing flow will be explained below.

[0255] Step 1:

[0256] Terminal: Collects real-time video from surveillance cameras and sensors installed in various locations. For example, a camera installed on a station platform captures video of the entire platform.

[0257] Step 2:

[0258] Terminal: Collected video data is temporarily stored in local storage, making it available for immediate analysis. For example, video captured by a camera is temporarily stored in a buffer.

[0259] Step 3:

[0260] Server: Performs filtering to remove noise from the received video data, for example, removing non-essential parts of the video data and converting it into a clear, easy-to-analyze format.

[0261] Step 4:

[0262] Server: Converts filtered video data to the optimal resolution and prepares it for sending to the analysis algorithm. Converts data into the format required by the analysis method.

[0263] Step 5:

[0264] Server: The filtered video data is fed into an AI model, where object and behavior recognition algorithms are applied, for example to detect people holding knives or handling suspicious liquids.

[0265] Step 6:

[0266] Server: Identifies specific abnormal behavior, such as people walking quickly or moving differently from the general crowd, thereby identifying suspicious individuals in real time.

[0267] Step 7:

[0268] Server: Runs an emotion engine based on video data and analyzes the user's facial expressions, voice, posture, etc. For example, it analyzes emotions such as smile, anger, and sadness from camera footage.

[0269] Step 8:

[0270] Server: Emotion data determined by the emotion engine is supplied to the analysis means and used as data to improve the accuracy of anomaly detection. For example, if a person with an angry expression is holding a knife, the degree of anomaly is set high.

[0271] Step 9:

[0272] Server: If an anomaly is detected based on the analysis results, a detailed report is generated, including the location, time, and type of anomaly detected.

[0273] Step 10:

[0274] Server: Transforms reports into a format that is easy to read and prepares them for notification, e.g., creating text and visual reports.

[0275] Step 11:

[0276] Server: Based on the generated report, it immediately notifies officials and police via email, SMS, in-app alerts, etc.

[0277] Step 12:

[0278] Terminal: A notification is sent to the employee's terminal and an alert sound is generated to attract attention. For example, a notification is displayed on the computer of an employee working in the monitoring room.

[0279] Step 13:

[0280] User: Upon receiving the notification, staff or police rush to the scene and check the details of the abnormality. For example, staff rush to the scene while checking the surveillance camera footage.

[0281] Step 14:

[0282] User: Eliminate the cause of the anomaly and take appropriate action to return the situation to safety. For example, staff will seize the dangerous material and coordinate with the police.

[0283] Step 15:

[0284] Server: Stores video footage of detected anomalies for later analysis or as evidence. For example, it stores video footage of before and after the anomaly occurred in a specific format.

[0285] Step 16:

[0286] Server: Analyzes the stored video data to evaluate and improve system performance, for example by retraining the AI ​​model based on past cases.

[0287] Example 2

[0288] 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."

[0289] Public places and educational institutions are required to quickly detect suspicious individuals and abnormal behavior and respond appropriately. However, conventional systems have low anomaly detection accuracy and frequently generate false alarms. Furthermore, they are unable to respond quickly in emergencies, making it difficult to prevent serious damage. Therefore, there is a need to develop a new system that improves the accuracy of anomaly detection and enables quick and appropriate responses.

[0290] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes an acquisition means for collecting data, a preprocessing means for filtering the acquired data and converting its resolution, and an emotion recognition means for recognizing the user's emotion based on the preprocessed data. This makes it possible to analyze abnormal behavior and the user's emotion with high accuracy and to take prompt and appropriate action.

[0291] An "acquisition means" is a device or group of devices that collects data to detect abnormalities.

[0292] The "analysis means" refers to a processing device or algorithm for detecting anomalies based on the acquired data.

[0293] The "notification means" refers to a device or software that notifies relevant parties such as staff and police of abnormalities detected by the analysis means.

[0294] The "pre-processing means" refers to a processing device or algorithm that filters the acquired image data and converts it to an optimal resolution.

[0295] "Emotion recognition means" refers to an algorithm or device for analyzing the user's emotions based on acquired video data.

[0296] An "object recognition algorithm" is a program or processing method for analyzing video data and identifying specific objects (e.g., knives, suspicious liquids).

[0297] A "behavior recognition algorithm" is a program or processing method for analyzing video data and detecting specific behaviors (e.g., suspicious movements).

[0298] The "emergency notification function" is a function that sends real-time notifications to staff and police when an abnormality is detected.

[0299] This invention is a system that quickly detects abnormalities in public places and educational institutions and automates appropriate responses. This system combines video acquisition means, analysis means, notification means, and an emotion engine that recognizes user emotions, significantly improving safety.

[0300] The overall configuration of the system includes an "acquisition means" that collects data to detect abnormalities, a "preprocessing means" that filters and converts the resolution of the acquired data, an "analysis means" that analyzes the preprocessed data, a "notification means" that notifies the user of the analysis results, and an "emotion recognition means" that recognizes the user's emotions.

[0301] Explanation of program processing

[0302] Below is a detailed description of the process at each stage of the system.

[0303] Data collection

[0304] Terminal: Surveillance cameras and sensors installed in various locations collect video in real time. For example, a camera installed on a station platform captures video of the entire platform. These videos are temporarily stored in local storage.

[0305] Terminal: Collected video data is stored in buffer memory so that it can be analyzed immediately.

[0306] Data Preprocessing

[0307] Server: Performs filtering to remove noise from the received video data. Removes digital noise and adjusts the contrast and brightness of the video.

[0308] Server: The filtered video data is converted to the optimal resolution and prepared for transmission to the analysis algorithm. Data is compressed and format converted.

[0309] Analysis for anomaly detection

[0310] Server: The filtered video data is fed into the AI ​​model and an object recognition algorithm is applied. For example, the AI ​​model identifies knives, suspicious liquids, etc.

[0311] Server: Then, behavioral recognition algorithms are used to detect certain anomalous behaviors (e.g., rapid movements or suspicious movements), and behavioral patterns are analyzed to identify behaviors that differ from normal behavior.

[0312] Analysis for Emotion Recognition

[0313] Server: Runs an emotion engine based on video data, analyzing the user's facial expressions, voice, and posture. For example, it can read emotions such as smile, anger, and sadness from camera footage.

[0314] Server: Emotion data determined by the emotion engine is supplied to the analysis means to improve the accuracy of anomaly detection. The emotion data adjusts the anomaly parameter, increasing the accuracy of abnormal behavior.

[0315] Identifying abnormal situations

[0316] Server: If an anomaly is detected based on the analysis results, a detailed report is generated, including the location, time, and type of anomaly.

[0317] Server: Converts reports into a user-friendly format and prepares them for notification. Automatically generates reports in text and visual formats.

[0318] Notification Processing

[0319] Server: Based on the generated report, it immediately notifies officials and police, specifically by launching a system to send emails, SMS, and in-app alerts.

[0320] Notification method: A notification is sent to the employee's device, an alert sound is heard to attract attention, and a pop-up notification is displayed on the PC screen of the employee in the monitoring room.

[0321] On-site response

[0322] User: Upon receiving the notification, staff and police rush to the scene to check the details of the abnormality. They head to the scene while checking the surveillance camera footage.

[0323] User: Eliminate the cause of the anomaly and take appropriate action to return the situation to safety. Staff will seize the dangerous object and take appropriate measures until the police arrive.

[0324] Video archiving and post-processing

[0325] Server: Stores video of detected anomalies for later analysis and as evidence. Stores video data in a secure storage and makes it accessible when needed.

[0326] Server: Uses the stored data to evaluate system performance and retrain the AI ​​model if necessary. Analyzes past anomaly detection cases and uses them as a dataset to improve the model's accuracy.

[0327] Specific examples

[0328] Specific example on a station platform

[0329] Terminal: Cameras installed on station platforms capture passenger activity in real time. Images of the platform during rush hour are stored in buffer memory.

[0330] Server: Receives video data and uses object and behavior recognition algorithms to detect people holding knives or making suspicious movements. The emotion engine then analyzes the person's emotions based on their facial expressions.

[0331] Server: Identifies a person holding a knife and with an angry expression as abnormal and generates a detailed report.

[0332] Notification method: If an abnormality is detected, an emergency notification will be sent to station staff and police via email and SMS. Staff will also receive a notification on their smartphones stating that "an angry person holding a knife is in the middle of the platform."

[0333] User: Staff rush to the scene and contact the police. Staff immediately head to the scene and handle the situation until the police arrive.

[0334] Example prompts to input to the generative AI model

[0335] "Analyze abnormal behavior and emotions based on video data collected by cameras installed on station platforms. Specifically, you need to detect people holding knives or behaving abnormally, and analyze their emotions as well."

[0336] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0337] Step 1:

[0338] Data collection

[0339] The terminal collects video in real time from surveillance cameras and sensors installed at various locations. Specifically, cameras installed on the station platform continuously capture video of the entire platform. (Input): Surveillance camera video data (Output): Acquired video data (stored in buffer memory)

[0340] The terminal stores the collected video data in buffer memory, making it available for immediate analysis.

[0341] Step 2:

[0342] Data Preprocessing

[0343] The server performs filtering on the video data received from the buffer memory to remove noise. Specifically, it removes digital noise from the video and adjusts the contrast and brightness. (Input): Video data in the buffer memory (Output): Video data after filtering

[0344] The server converts the filtered video data to the optimal resolution and prepares it for transmission to the analysis algorithm. Specific operations include data compression and format conversion. (Input): Filtered video data (Output): Resolution-converted video data

[0345] Step 3:

[0346] object recognition

[0347] The server inputs the resolution-converted video data into the AI ​​model and applies an object recognition algorithm. Specifically, the AI ​​model identifies knives, suspicious liquids, etc. (Input): Resolution-converted video data (Output): Object recognition results (data identifying knives, suspicious objects, etc.)

[0348] Step 4:

[0349] behavioral recognition

[0350] The server uses a behavior recognition algorithm based on the results of object recognition to detect specific abnormal behavior. Specifically, the AI ​​model analyzes behavior patterns and identifies behaviors that differ from normal behavior. (Input): Object recognition results (Output): Behavior recognition results (data identifying abnormal behavior)

[0351] Step 5:

[0352] emotion recognition

[0353] The server runs an emotion engine based on the behavioral recognition data, analyzing the user's facial expressions, voice, and posture. Specific behaviors include smiles, anger, sadness, and other emotions read from the camera footage. (Input): Behavior recognition results (Output): Emotion recognition results (data with identified emotions)

[0354] The server supplies the emotion data determined by the emotion engine to the analysis means to improve the accuracy of anomaly detection. (Input): Emotion recognition results (Output): Data with anomaly level parameters adjusted

[0355] Step 6:

[0356] Anomaly detection and report generation

[0357] If the server determines that an anomaly has occurred based on the analysis results, it generates a detailed report. Specifically, it creates a report that clearly indicates the location, time, and type of anomaly. (Input): Data with anomaly parameter adjustments (Output): Anomaly detection report

[0358] The server converts the generated report into an easy-to-read format and prepares it for notification. Specific operations include automatically generating reports in text and visual formats. (Input): Anomaly detection report (Output): Report that can be notified

[0359] Step 7:

[0360] Notification Processing

[0361] The server immediately notifies officials and police based on the contents of the notifiable report. Specific actions include launching a system to send emails, SMS, and in-app alerts. (Input): Notifiable report (Output): Sent notification (email, SMS, etc.)

[0362] The notification method is that a notification is sent to the employee's terminal and an alert sound is emitted to attract their attention. Specific actions include a pop-up notification being displayed on the PC screen in the monitoring room. (Input): Sent notification (Output): Notification displayed on the employee's terminal

[0363] Step 8:

[0364] On-site response

[0365] Upon receiving the notification, the staff or police rush to the scene and check the details of the abnormality. Specific actions include heading to the scene while checking the video from the surveillance camera. (Input): Notification displayed on staff's device (Output): Response begins at the scene

[0366] The user must take appropriate action to remove the cause of the anomaly and return the situation to safety. Specifically, the user must take appropriate action until the staff seizes the dangerous object and the police arrive. (Input): Start of on-site response (Output): Completion of anomaly response

[0367] Step 9:

[0368] Video archiving and post-processing

[0369] The server stores video of detected anomalies for later analysis and as evidence. Specific operations include storing the video data in a secure storage and making it accessible as needed. (Input): On-site response data (Output): Stored video of anomalies

[0370] The server uses the stored data to evaluate the system's performance and retrain the AI ​​model as necessary. Specifically, it analyzes past anomaly detection cases and uses them as a dataset to improve the model's accuracy. (Input): Stored anomaly video. (Output): Updated AI model.

[0371] (Application example 2)

[0372] 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."

[0373] Modern society requires rapid and appropriate responses to various abnormal situations that may occur in public places and educational institutions (for example, the appearance of a person armed with a knife or the abandonment of suspicious liquids). However, previous surveillance systems have focused on object recognition and behavioral recognition, making it difficult to combine emotion recognition for highly accurate anomaly detection. Furthermore, real-time notification capabilities have been limited, resulting in delayed responses when an abnormality occurs. This has sometimes led to insufficient safety in public places and educational institutions.

[0374] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0375] In this invention, the server includes a data acquisition means for collecting visual data for detecting anomalies, a data analysis means for detecting anomalies from the acquired visual data, a notification means for notifying anomalies detected by the data analysis means, an emotion recognition means for recognizing a user's emotion, and an integrated analysis means for determining a detection result by combining abnormal behavior and the user's emotion, thereby enabling highly accurate detection of anomalies and rapid response.

[0376] "Visual data" is digital data containing visual information that is collected by image capture devices such as cameras and sensors.

[0377] "Data acquisition means" is a general term for devices including cameras and various sensors installed to collect visual data.

[0378] "Data analysis means" includes software and hardware means for analyzing acquired visual data and detecting specific patterns or anomalies.

[0379] "Notification Methods" are the functions and devices used to notify personnel or law enforcement in real time of details of detected anomalies, including notifications in the form of email, SMS, in-app alerts, etc.

[0380] "Emotion recognition means" refers to algorithms and devices that analyze a user's facial expressions, voice, posture, etc., and identify their psychological state and emotions.

[0381] The "integrated analysis means" is a means for combining abnormal behavior and user emotion data to perform a comprehensive analysis and improve the accuracy of abnormality detection.

[0382] An "object recognition algorithm" is a computational procedure for detecting specific objects (e.g., knives, suspicious liquids) from video data.

[0383] A "behavior recognition algorithm" is a calculation procedure for analyzing the movements and actions of people in video data and identifying suspicious behavior.

[0384] The "storage means" is a device for storing video data of detected abnormalities for later analysis or use as evidence.

[0385] This invention is a system that quickly detects abnormalities and automates appropriate responses, improving safety in public places and educational institutions. This system is composed of a combination of data acquisition means, data analysis means, reporting means, emotion recognition means, and integrated analysis means.

[0386] Data Acquisition Method

[0387] The devices use cameras and sensors installed at each location to collect visual data, providing real-time footage of the entire public space or educational institution. For example, a camera installed on a train platform captures footage of the entire platform.

[0388] Data Analysis Methods

[0389] The server analyzes the acquired visual data. This analysis includes object recognition algorithms and behavior recognition algorithms. For example, it detects people holding knives or handling suspicious liquids. These object recognition and behavior recognition algorithms are implemented using frameworks such as OpenCV and Keras.

[0390] emotion recognition means

[0391] The server is equipped with emotion recognition means to analyze the user's facial expressions, voice, and posture, allowing it to identify a person's emotional state (e.g., anger, fear, sadness, etc.). The technologies used include deep learning models and supporting software such as TENSORFLOW® and Keras models.

[0392] Integrated analysis method

[0393] The server combines abnormal behavior and user emotional data for comprehensive analysis, improving the accuracy of anomaly detection. This integrated analysis method further improves the accuracy of anomaly detection when, for example, the user is not only holding a knife but also has an angry expression.

[0394] Reporting means

[0395] The server provides a means of notification to notify personnel and police in real time of detected anomalies, including details of the anomalies, via email, SMS, in-app alerts, etc. Video data of the anomalies is stored in the cloud or local storage for later analysis or as evidence.

[0396] Specific examples

[0397] Cameras installed on station platforms capture passenger activity in real time. For example, the cameras capture a situation where there are many passengers on the platform during rush hour. The server receives the video data and first uses object recognition and behavior recognition algorithms to detect people holding knives or people behaving suspiciously. The server then analyzes the person's emotions from facial expressions captured by emotion recognition means. If a person holding a knife or with an angry expression is judged to be abnormal, an emergency notification is immediately sent to staff and the police. Staff receive the notification and rush to the scene.

[0398] Prompt Sentence Examples

[0399] It analyzes camera footage in real time to detect suspicious individuals and movements. It also uses facial recognition technology to analyze emotions and immediately notify staff and police if an abnormality is detected. Please list the hardware and software required for this system and explain how to implement it specifically.

[0400] This system is expected to enable rapid and highly accurate detection of abnormalities in public places and educational institutions, significantly improving safety.

[0401] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0402] Step 1:

[0403] The device uses cameras and sensors installed at each location to collect visual data. The cameras and sensors capture video data in real time and temporarily store the data in local storage. The input is the visual data from the camera, and the output is the temporarily stored video data.

[0404] Step 2:

[0405] The server preprocesses the video data received from the terminal. First, it filters the data to remove noise, and then converts it to the required resolution. This process prepares the data for the data analysis means to operate efficiently. The input is the temporarily stored video data, and the output is the filtered and resolution-converted video data.

[0406] Step 3:

[0407] The server analyzes the filtered video data. In this step, it applies object and behavior recognition algorithms to detect abnormal behaviors and objects. For example, it identifies people holding knives or suspicious liquids. The input is the filtered and resolution-converted video data, and the output is the detection result data indicating anomalies.

[0408] Step 4:

[0409] The server analyzes the user's emotions based on the detection result data. Using emotion recognition means, it analyzes the user's facial expressions, voice, posture, etc., and identifies emotions such as anger, sadness, and fear. The input is the detection result data and visual data regarding abnormal behavior and abnormal objects, and the output is the emotion recognition result data.

[0410] Step 5:

[0411] The server integrates the abnormal behavior and emotion recognition results to make a final anomaly determination. Specifically, if a person holding a knife also displays anger, it is judged to be highly abnormal. The input is abnormal behavior data and emotion recognition result data, and the output is the integrated anomaly detection result.

[0412] Step 6:

[0413] The server activates the notification mechanism based on the final anomaly detection results. Emergency notifications are sent to personnel and police in real time, and detailed reports are generated. Notification formats include email, SMS, and in-app alerts, and messages containing anomaly identification information are sent. The input is the consolidated anomaly detection results, and the output is emergency notifications and detailed reports.

[0414] Step 7:

[0415] The user receives notifications from the reporting means and responds promptly to the scene. Specifically, when a staff member receives a notification, he or she immediately heads to the scene to confirm and respond to the abnormality. The input is the emergency notification and detailed report, and the output is the actual abnormality response action.

[0416] Step 8:

[0417] The server stores video data when an anomaly is detected. The stored data is used for later analysis and as evidence. The input is the video data when an anomaly occurs, and the output is the data stored in the cloud or local storage.

[0418] This series of processing steps enables the system to detect anomalies with high accuracy and respond quickly, significantly improving safety in public places and educational institutions.

[0419] 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.

[0420] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (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.

[0421] 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.

[0422] [Second embodiment]

[0423] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0424] 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.

[0425] 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).

[0426] 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.

[0427] 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.

[0428] 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).

[0429] 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.

[0430] 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.

[0431] 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.

[0432] 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.

[0433] 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.

[0434] 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."

[0435] This invention is a system that quickly detects abnormalities in public places and educational institutions and automates appropriate responses. This system combines video capture, analysis, and notification means to significantly improve safety at the site.

[0436] Program processing

[0437] 1. Data Collection

[0438] Terminal: Collects real-time video from surveillance cameras and sensors installed in various locations. For example, a camera installed on a station platform captures video of the entire platform.

[0439] Terminal: Temporarily stores collected video data and makes it available for immediate analysis. For example, video captured by a camera is temporarily stored in local storage.

[0440] 2. Data Preprocessing

[0441] Server: Filters the received video data to remove noise, converting it into a clear, easy-to-analyze image.

[0442] Server: Converts the filtered video data to the optimal resolution and prepares it for sending to the analysis means.

[0443] 3. Analysis for Anomaly Detection

[0444] Server: The filtered video data is fed into an AI model, where object and behavior recognition algorithms are applied, for example to detect people holding knives or handling suspicious liquids.

[0445] Server: Identifies specific abnormal behavior, such as detecting someone walking quickly or moving differently from the normal crowd.

[0446] 4. Identifying abnormal situations

[0447] Server: If an anomaly is detected based on the analysis results, a detailed report is generated, including the location, time, and type of anomaly detected.

[0448] Server: Transforms the report into a readable format and prepares it for notification.

[0449] 5. Notification Processing

[0450] Server: Based on the generated report, it immediately notifies officials and police via email, SMS, in-app alerts, etc.

[0451] Notification method: A notification is sent to the employee's device and an alert sound is generated to attract attention. For example, a notification is displayed on the computer of an employee working in the monitoring room.

[0452] 6. On-site response

[0453] User: Upon receiving the notification, staff or police rush to the scene and check the details of the abnormality. For example, staff rush to the scene while checking the surveillance camera footage.

[0454] User: Eliminate the cause of the anomaly and take appropriate action to return the situation to safety.

[0455] 7. Video storage and post-processing

[0456] Server: Stores video footage of detected anomalies for later analysis or as evidence. For example, it stores video footage of before and after the anomaly occurred in a specific format.

[0457] Server: Analyzes the stored video data and uses it to evaluate and improve system performance.

[0458] Specific examples

[0459] Example on a station platform:

[0460] Terminal: Cameras installed on station platforms capture passenger activity in real time. For example, cameras can capture the situation of large numbers of passengers on the platform during rush hour.

[0461] Server: Receives video data and analyzes abnormal behavior, such as detecting someone pulling out a knife in a crowd.

[0462] Notification method: If an abnormality is detected, an emergency notification will be sent to station staff and the police via email or SMS. For example, a staff member may receive a notification on their smartphone stating that "a person with a knife is in the center of the platform."

[0463] User: Staff rush to the scene and contact the police. For example, staff immediately head to the scene and handle the situation until the police arrive.

[0464] Thus, the present invention is a system that provides an embodiment for quickly detecting abnormalities in public places and educational institutions and taking appropriate measures.

[0465] The processing flow will be explained below.

[0466] Step 1:

[0467] Terminal: Collects real-time video from surveillance cameras and sensors installed in various locations. For example, a camera installed on a station platform captures video of the entire platform.

[0468] Step 2:

[0469] Terminal: Collected video data is temporarily stored in local storage, making it available for immediate analysis. For example, video captured by a camera is temporarily stored in a buffer.

[0470] Step 3:

[0471] Server: Performs filtering to remove noise from the received video data, for example, removing non-essential parts of the video data and converting it into a clear, easy-to-analyze format.

[0472] Step 4:

[0473] Server: Converts filtered video data to the optimal resolution and prepares it for sending to the analysis algorithm. Converts data into the format required by the analysis method.

[0474] Step 5:

[0475] Server: The filtered video data is fed into an AI model, where object and behavior recognition algorithms are applied, for example to detect people holding knives or handling suspicious liquids.

[0476] Step 6:

[0477] Server: Identifies specific abnormal behavior, such as people walking quickly or moving differently from the general crowd, thereby identifying suspicious individuals in real time.

[0478] Step 7:

[0479] Server: If an anomaly is detected based on the analysis results, a detailed report is generated, including the location, time, and type of anomaly detected.

[0480] Step 8:

[0481] Server: Transforms reports into a format that is easy to read and prepares them for notification, e.g., creating text and visual reports.

[0482] Step 9:

[0483] Server: Based on the generated report, it immediately notifies officials and police via email, SMS, in-app alerts, etc.

[0484] Step 10:

[0485] Terminal: A notification is sent to the employee's terminal and an alert sound is generated to attract attention. For example, a notification is displayed on the computer of an employee working in the monitoring room.

[0486] Step 11:

[0487] User: Upon receiving the notification, staff or police rush to the scene and check the details of the abnormality. For example, staff rush to the scene while checking the surveillance camera footage.

[0488] Step 12:

[0489] User: Eliminate the cause of the anomaly and take appropriate action to return the situation to safety. For example, staff will seize the dangerous material and coordinate with the police.

[0490] Step 13:

[0491] Server: Stores video footage of detected anomalies for later analysis or as evidence. For example, it stores video footage of before and after the anomaly occurred in a specific format.

[0492] Step 14:

[0493] Server: Analyzes the stored video data to evaluate and improve system performance, for example by retraining the AI ​​model based on past cases.

[0494] Example 1

[0495] 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."

[0496] Public places and educational institutions require the ability to quickly and accurately detect anomalies and automate appropriate responses. However, conventional systems lack the accuracy of anomaly detection and notification speed required, making manual monitoring and response necessary. Furthermore, the process of storing detected anomalies and post-analysis was often inefficient. A system that solves these problems and significantly improves safety is needed.

[0497] 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.

[0498] In this invention, the server includes a data acquisition means for collecting video data for detecting anomalies, a data preprocessing means for removing noise from the acquired video data and optimizing resolution, an analysis means for detecting anomalies from the video data processed by the data preprocessing means, a notification means for notifying anomalies detected by the analysis means, and a video storage means for storing video of the detected anomalies for use as later analysis or evidence. This enables rapid and accurate detection of anomalies, real-time notification and storage, and further subsequent analysis.

[0499] "Data acquisition means" refers to the devices and technologies that collect the video data necessary to detect abnormalities.

[0500] "Data pre-processing means" refers to technology or devices that perform processing to remove noise from acquired video data and optimize resolution.

[0501] "Analysis means" refers to a device that includes algorithms and techniques for detecting anomalies based on pre-processed video data.

[0502] "Notification means" refers to the technology or device used to notify relevant parties of abnormalities detected by the analysis means.

[0503] "Video storage means" refers to technology or equipment for storing video of detected abnormalities and using it for later analysis or as evidence.

[0504] "Object recognition algorithm" refers to the computational processes and techniques used to detect and identify objects based on video data.

[0505] "Behavior recognition algorithm" refers to the computational processes and technologies used to detect and identify behaviors based on video data and determine abnormal behavior.

[0506] "Emergency notification function" refers to technology and systems that notify relevant parties such as staff and police in real time of details of detected abnormalities.

[0507] The present invention relates to a system for quickly detecting abnormalities in public places and educational institutions and automating appropriate responses. The system includes a data acquisition unit, a data preprocessing unit, an analysis unit, a notification unit, and a video storage unit.

[0508] The system is programmed as follows:

[0509] First, surveillance cameras and other data acquisition methods installed at various locations collect video data in real time. For example, cameras installed on train platforms capture passenger movements.

[0510] The device then temporarily stores the collected video data and makes it available for immediate analysis using local storage.

[0511] The server then performs a filtering process to remove noise from the received video data, making the video clearer and easier to analyze. The filtered video data is then converted to an optimal resolution and prepared for transmission to the analysis means.

[0512] The analysis method inputs the filtered video data into an AI model and applies object and behavior recognition algorithms to detect, for example, people holding knives or handling suspicious liquids, as well as identify specific abnormal behaviors, such as people walking quickly or moving differently from the general crowd.

[0513] If the analysis results in an anomaly, the server generates a detailed report that includes the location, time, and type of anomaly detected, and converts the report into an easy-to-read format for notification.

[0514] The notification method will immediately notify employees and police based on the contents of the generated report. Notifications will be sent via email, SMS, in-app alerts, etc. For example, an emergency notification such as "There is a person with a knife on the station platform" will be sent to an employee's smartphone.

[0515] The users who receive the notification, i.e., officials or police, rush to the scene and check the details of the abnormality. Officials rush to the scene while checking the surveillance camera footage and take appropriate action to eliminate the cause of the abnormality.

[0516] Finally, the server stores the video of the detected anomaly for later analysis and as evidence. The stored video data can be analyzed to evaluate and improve the system's performance.

[0517] Prompt Sentence Examples

[0518] Below are some example prompts to input to the generative AI model:

[0519] "Please explain the most effective method for detecting abnormal behavior in public places."

[0520] "What are the key elements of a system that quickly detects anomalies and automatically notifies you?"

[0521] Thus, the present invention is a system that provides an embodiment for quickly detecting abnormalities in public places and educational institutions and taking appropriate measures.

[0522] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0523] Step 1: Data collection

[0524] Terminal: Surveillance cameras installed in various locations collect video data in real time. For example, cameras installed on train platforms capture passenger movements.

[0525] Input: Real-time video data from public places and educational institutions.

[0526] Output: Real-time video data.

[0527] Step 2: Temporarily save the video data

[0528] Terminal: Collected video data is temporarily stored in local storage. For example, video data captured by a camera is temporarily stored on a local drive so that it can be analyzed immediately.

[0529] Input: Real-time video data.

[0530] Output: Stored temporary video data.

[0531] Step 3: Filtering

[0532] Server: Filters the received video data to remove noise. By removing unnecessary noise and blur, the video becomes clear and easier to analyze.

[0533] Input: Stored temporary video data.

[0534] Output: Filtered video data.

[0535] Step 4: Optimize the resolution

[0536] Server: Converts the filtered video data to the appropriate resolution and prepares it for analysis. For example, compresses high-resolution video slightly to maintain analysis accuracy and speed.

[0537] Input: Filtered video data.

[0538] Output: Optimized resolution video data.

[0539] Step 5: Input to the AI ​​model

[0540] Server: The optimized resolution video data is input into the AI ​​model and an object recognition algorithm is applied to detect people and objects in the video data.

[0541] Input: Video data with optimized resolution.

[0542] Output: Object and person detection results.

[0543] Step 6: Applying the activity recognition algorithm

[0544] Server: Uses the results of the object recognition algorithm to apply the behavior recognition algorithm, for example, to detect people whose movements deviate from their normal behavioral patterns.

[0545] Input: Object and person detection results.

[0546] Output: Abnormal behavior detection results.

[0547] Step 7: Generate anomaly detection reports

[0548] Server: If the analysis results in an anomaly, a detailed report is generated, including the location, time, and specific details of the anomaly.

[0549] Input: Anomalous behavior detection results.

[0550] Output: Anomaly detection report.

[0551] Step 8: Convert the report to another format

[0552] Server: Converts the generated reports into a format that is easy to read, for example, creating reports in PDF or HTML format.

[0553] Input: Anomaly detection report.

[0554] Output: Formatted report.

[0555] Step 9: Call Emergency Services

[0556] Server: Based on the generated reports, it provides instant notifications to officials and police via email, SMS or in-app alerts.

[0557] Input: The formatted report.

[0558] Output: Emergency call notification.

[0559] Step 10: Displaying notifications

[0560] Notification method: An emergency notification will be displayed on the employee's terminal and an alert sound will be emitted to attract attention. For example, the notification will be displayed immediately on the computer of an employee on duty in the monitoring room.

[0561] Input: Emergency call notification.

[0562] Output: Notifications that appear on the device.

[0563] Step 11: Rush to the scene

[0564] User: Upon receiving the notification, staff and police rush to the scene to check the details of the abnormality. Staff head to the scene while checking the surveillance camera footage.

[0565] Input: Notifications that appear on your device.

[0566] Output: Check the details of the error.

[0567] Step 12: Dealing with abnormalities

[0568] User: Eliminate the cause of the anomaly and take appropriate action to return the situation to safety. For example, staff may apprehend the anomaly.

[0569] Input: Check the details of the abnormality.

[0570] Output: Safe state restored.

[0571] Step 13: Save the abnormal video

[0572] Server: Stores video footage of detected anomalies for later analysis and as evidence. Stores footage of several minutes before and after the anomaly occurs.

[0573] Input: Detailed confirmation of anomaly and video data.

[0574] Output: Archived footage of the anomaly.

[0575] Step 14: Performance evaluation and system improvement

[0576] Server: Analyzes the stored video data to evaluate and improve the system's performance. For example, it evaluates detection accuracy and notification speed, and improves the AI ​​model.

[0577] Input: Archived anomalous footage.

[0578] Output: System improvement measures and evaluation results.

[0579] (Application example 1)

[0580] 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."

[0581] In modern society, there is a demand for immediate detection and rapid response of abnormal behavior in public places and educational institutions. However, current systems can be slow to detect and notify abnormalities, or on-site staff may not be able to respond immediately. Furthermore, there is a lack of means to immediately communicate detailed information about abnormalities to staff visually and audibly, making it difficult to respond to abnormalities efficiently. The objective of the present invention is to solve these problems.

[0582] 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.

[0583] In this invention, the server includes a video acquisition means for collecting camera data for detecting abnormalities, an analysis means for detecting abnormalities from the acquired camera data, a notification means for notifying the user of the abnormality detected by the analysis means, and a wearable display means for immediately conveying the notification information from the notification means to the user's visual and auditory senses. This makes it possible to quickly detect abnormal behavior, notify staff in real time, and respond immediately on site.

[0584] "Video acquisition means" refers to equipment and software that collects camera data for detecting abnormalities.

[0585] "Analysis means" refers to algorithms and software for detecting anomalies in acquired camera data.

[0586] The "notification means" refers to a system and software for notifying staff and other relevant parties of abnormalities detected by the analysis means.

[0587] The "wearable display means" is a device for instantly conveying the notification information of the notification means to the user's visual and auditory senses.

[0588] An "object recognition algorithm" is a program for identifying specific objects (e.g., knives, suspicious liquids) from video data.

[0589] A "behavior recognition algorithm" is a program that analyzes people's movements from video data and detects abnormal behavior (e.g., suspicious movements).

[0590] The "emergency notification function" is a function that notifies staff and police in real time of detected abnormalities.

[0591] A "server" is a main computing device that performs a series of processes such as data collection, analysis, and notification.

[0592] "User" refers to the staff and other relevant parties who receive and respond to the notified abnormality information.

[0593] The system for realizing this invention is configured using the following hardware and software: The hardware used is a surveillance camera, a cloud server, and smart glasses, while the software used is an AI model for video analysis (e.g., YOLO, OpenPose), a library for video noise reduction and filtering (e.g., OpenCV), and a notification service API (e.g., Firebase Cloud Messaging, Twilio).

[0594] The server collects video data from surveillance cameras in real time, which is first pre-processed with a program to remove noise and perform filtering, improving the quality of the video data and making it easier to analyze.

[0595] The server then inputs the preprocessed video data into an AI model and applies object and behavior recognition algorithms, such as YOLO to detect knives and suspicious liquids and OpenPose to identify abnormal behaviors, enabling the detection of specific abnormal behaviors and objects in real time.

[0596] If the server detects an anomaly based on these analysis results, it generates a detailed report containing information such as the location, time, and type of anomaly. After the report is generated, an emergency notification is sent immediately to staff and relevant parties via the notification service API. Notifications are sent in multiple ways, including email, SMS, and in-app alerts.

[0597] The user, a staff member, receives notifications through the smart glasses. The wearable display immediately conveys the notification information visually and audibly, allowing the staff member to quickly grasp the situation on-site. The staff member can check the details of any abnormalities in real time and take prompt action.

[0598] Examples:

[0599] For example, if a surveillance camera installed on a station platform detects a person holding a knife in the crowd, the server will filter the footage and identify the person as holding a knife using the YOLO model. If an anomaly is detected, the server will generate a detailed report and send a notification to the employee's smart glasses via the notification service API, stating, "There is a person holding a knife in the center of the platform." The employee will receive the notification and confirm the visual and audio alert, allowing them to respond to the scene quickly.

[0600] Example prompt sentence:

[0601] "Analyze the footage from the center of the platform to detect if there is anyone holding a knife."

[0602] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0603] Step 1:

[0604] The device transmits real-time video data captured by the camera to the cloud server. The input is video data from the surveillance camera, and the output is temporarily stored video data. This process involves acquiring video data and immediately transferring it to the server.

[0605] Step 2:

[0606] The server performs filtering to remove noise from the received video data. The input is the temporarily stored video data, and the output is the noise-removed video data. This process uses the OpenCV library to process the data to improve the video quality.

[0607] Step 3:

[0608] The server converts the filtered video data to the optimal resolution and prepares it for transmission to the analysis means. The input is the noise-removed video data, and the output is video data converted to a resolution suitable for analysis. This process involves data calculations using high-performance image processing algorithms.

[0609] Step 4:

[0610] The server then inputs the converted video data into an AI model and applies object and action recognition algorithms. The input is video data suitable for analysis, and the output is detected anomaly information. This process uses YOLO and OpenPose for object and action recognition and data calculation.

[0611] Step 5:

[0612] The server determines anomalies based on the analysis results of the AI ​​model and generates a detailed report. The input is the detected anomaly information, and the output is a report summarizing the details of the anomaly. This process generates a document summarizing the type of anomaly, the location where it occurred, the time, etc.

[0613] Step 6:

[0614] Based on the generated report, the server sends emergency notifications to staff and related parties via the notification service API. The input is a detailed report, and the output is the sent notification. This process uses services such as Firebase Cloud Messaging and Twilio to send notifications immediately.

[0615] Step 7:

[0616] The user receives notifications through smart glasses and acknowledges visual and audible alerts. The input is the notification sent, and the output is the user's acknowledgement and immediate response. This process uses the wearable device to display information and provide alerts.

[0617] Step 8:

[0618] After checking the contents of the notification, the user quickly heads to the site, checks the details of the abnormality, and takes action. The input is the content of the notification and the situation at the site, and the output is the appropriate action to eliminate the cause of the abnormality. This process allows for a quick response and problem resolution at the site.

[0619] 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.

[0620] This invention is a system that quickly detects abnormalities in public places and educational institutions and automates appropriate responses. This system combines video acquisition means, analysis means, notification means, and an emotion engine that recognizes user emotions, significantly improving safety.

[0621] Program processing

[0622] 1. Data Collection

[0623] Terminal: Collects real-time video from surveillance cameras and sensors installed in various locations. For example, a camera installed on a station platform captures video of the entire platform.

[0624] Terminal: Collected video data is temporarily stored in local storage, making it available for immediate analysis. For example, video captured by a camera is temporarily stored in a buffer.

[0625] 2. Data Preprocessing

[0626] Server: Performs filtering to remove noise from the received video data, for example, removing non-essential parts of the video data and converting it into a clear, easy-to-analyze format.

[0627] Server: Converts filtered video data to the optimal resolution and prepares it for sending to the analysis algorithm. Converts data into the format required by the analysis method.

[0628] 3. Analysis for Anomaly Detection

[0629] Server: The filtered video data is fed into an AI model, where object and behavior recognition algorithms are applied, for example to detect people holding knives or handling suspicious liquids.

[0630] Server: Identifies specific abnormal behavior, such as people walking quickly or moving differently from the general crowd, thereby identifying suspicious individuals in real time.

[0631] 4. Analysis for Emotion Recognition

[0632] Server: Runs an emotion engine based on video data and analyzes the user's facial expressions, voice, posture, etc. For example, it analyzes emotions such as smile, anger, and sadness from camera footage.

[0633] Server: Emotion data determined by the emotion engine is supplied to the analysis means and used as data to improve the accuracy of anomaly detection. For example, if a person with an angry expression is holding a knife, the degree of anomaly is set high.

[0634] 5. Identifying abnormal situations

[0635] Server: If an anomaly is detected based on the analysis results, a detailed report is generated, including the location, time, and type of anomaly detected.

[0636] Server: Transforms reports into a format that is easy to read and prepares them for notification, e.g., creating text and visual reports.

[0637] 6. Notification Processing

[0638] Server: Based on the generated report, it immediately notifies officials and police via email, SMS, in-app alerts, etc.

[0639] Notification method: A notification is sent to the employee's device and an alert sound is generated to attract attention. For example, a notification is displayed on the computer of an employee working in the monitoring room.

[0640] 7. On-site response

[0641] User: Upon receiving the notification, staff or police rush to the scene and check the details of the abnormality. For example, staff rush to the scene while checking the surveillance camera footage.

[0642] User: Eliminate the cause of the anomaly and take appropriate action to return the situation to safety. For example, staff will seize the dangerous material and coordinate with the police.

[0643] 8. Video storage and post-processing

[0644] Server: Stores video footage of detected anomalies for later analysis or as evidence. For example, it stores video footage of before and after the anomaly occurred in a specific format.

[0645] Server: Analyzes the stored video data to evaluate and improve system performance, for example by retraining the AI ​​model based on past cases.

[0646] Specific examples

[0647] Example on a station platform:

[0648] Terminal: Cameras installed on station platforms capture passenger activity in real time. For example, cameras can capture the situation of large numbers of passengers on the platform during rush hour.

[0649] Server: After receiving the video data, the server first uses object recognition and behavior recognition algorithms to detect people holding a knife or making suspicious movements.Then, the server analyzes the person's emotions based on facial expressions captured by the emotion engine.

[0650] Server: A person holding a knife and with an angry expression is judged to be abnormal.

[0651] Notification method: If an abnormality is detected, an emergency notification will be sent to station staff and the police via email or SMS. For example, a staff member may receive a notification on their smartphone saying, "There is an angry person holding a knife in the middle of the platform."

[0652] User: Staff rush to the scene and contact the police. For example, staff immediately head to the scene and handle the situation until the police arrive.

[0653] In this way, by combining emotion engines, the present invention is a system that provides an embodiment for detecting anomalies with higher accuracy and taking prompt and appropriate action.

[0654] The processing flow will be explained below.

[0655] Step 1:

[0656] Terminal: Collects real-time video from surveillance cameras and sensors installed in various locations. For example, a camera installed on a station platform captures video of the entire platform.

[0657] Step 2:

[0658] Terminal: Collected video data is temporarily stored in local storage, making it available for immediate analysis. For example, video captured by a camera is temporarily stored in a buffer.

[0659] Step 3:

[0660] Server: Performs filtering to remove noise from the received video data, for example, removing non-essential parts of the video data and converting it into a clear, easy-to-analyze format.

[0661] Step 4:

[0662] Server: Converts filtered video data to the optimal resolution and prepares it for sending to the analysis algorithm. Converts data into the format required by the analysis method.

[0663] Step 5:

[0664] Server: The filtered video data is fed into an AI model, where object and behavior recognition algorithms are applied, for example to detect people holding knives or handling suspicious liquids.

[0665] Step 6:

[0666] Server: Identifies specific abnormal behavior, such as people walking quickly or moving differently from the general crowd, thereby identifying suspicious individuals in real time.

[0667] Step 7:

[0668] Server: Runs an emotion engine based on video data and analyzes the user's facial expressions, voice, posture, etc. For example, it analyzes emotions such as smile, anger, and sadness from camera footage.

[0669] Step 8:

[0670] Server: Emotion data determined by the emotion engine is supplied to the analysis means and used as data to improve the accuracy of anomaly detection. For example, if a person with an angry expression is holding a knife, the degree of anomaly is set high.

[0671] Step 9:

[0672] Server: If an anomaly is detected based on the analysis results, a detailed report is generated, including the location, time, and type of anomaly detected.

[0673] Step 10:

[0674] Server: Transforms reports into a format that is easy to read and prepares them for notification, e.g., creating text and visual reports.

[0675] Step 11:

[0676] Server: Based on the generated report, it immediately notifies officials and police via email, SMS, in-app alerts, etc.

[0677] Step 12:

[0678] Terminal: A notification is sent to the employee's terminal and an alert sound is generated to attract attention. For example, a notification is displayed on the computer of an employee working in the monitoring room.

[0679] Step 13:

[0680] User: Upon receiving the notification, staff or police rush to the scene and check the details of the abnormality. For example, staff rush to the scene while checking the surveillance camera footage.

[0681] Step 14:

[0682] User: Eliminate the cause of the anomaly and take appropriate action to return the situation to safety. For example, staff will seize the dangerous material and coordinate with the police.

[0683] Step 15:

[0684] Server: Stores video footage of detected anomalies for later analysis or as evidence. For example, it stores video footage of before and after the anomaly occurred in a specific format.

[0685] Step 16:

[0686] Server: Analyzes the stored video data to evaluate and improve system performance, for example by retraining the AI ​​model based on past cases.

[0687] Example 2

[0688] 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."

[0689] Public places and educational institutions are required to quickly detect suspicious individuals and abnormal behavior and respond appropriately. However, conventional systems have low anomaly detection accuracy and frequently generate false alarms. Furthermore, they are unable to respond quickly in emergencies, making it difficult to prevent serious damage. Therefore, there is a need to develop a new system that improves the accuracy of anomaly detection and enables quick and appropriate responses.

[0690] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes an acquisition means for collecting data, a preprocessing means for filtering the acquired data and converting its resolution, and an emotion recognition means for recognizing the user's emotion based on the preprocessed data. This makes it possible to analyze abnormal behavior and the user's emotion with high accuracy and to take prompt and appropriate action.

[0691] An "acquisition means" is a device or group of devices that collects data to detect abnormalities.

[0692] The "analysis means" refers to a processing device or algorithm for detecting anomalies based on the acquired data.

[0693] The "notification means" refers to a device or software that notifies relevant parties such as staff and police of abnormalities detected by the analysis means.

[0694] The "pre-processing means" refers to a processing device or algorithm that filters the acquired image data and converts it to an optimal resolution.

[0695] "Emotion recognition means" refers to an algorithm or device for analyzing the user's emotions based on acquired video data.

[0696] An "object recognition algorithm" is a program or processing method for analyzing video data and identifying specific objects (e.g., knives, suspicious liquids).

[0697] A "behavior recognition algorithm" is a program or processing method for analyzing video data and detecting specific behaviors (e.g., suspicious movements).

[0698] The "emergency notification function" is a function that sends real-time notifications to staff and police when an abnormality is detected.

[0699] This invention is a system that quickly detects abnormalities in public places and educational institutions and automates appropriate responses. This system combines video acquisition means, analysis means, notification means, and an emotion engine that recognizes user emotions, significantly improving safety.

[0700] The overall configuration of the system includes an "acquisition means" that collects data to detect abnormalities, a "preprocessing means" that filters and converts the resolution of the acquired data, an "analysis means" that analyzes the preprocessed data, a "notification means" that notifies the user of the analysis results, and an "emotion recognition means" that recognizes the user's emotions.

[0701] Explanation of program processing

[0702] Below is a detailed description of the process at each stage of the system.

[0703] Data collection

[0704] Terminal: Surveillance cameras and sensors installed in various locations collect video in real time. For example, a camera installed on a station platform captures video of the entire platform. These videos are temporarily stored in local storage.

[0705] Terminal: Collected video data is stored in buffer memory so that it can be analyzed immediately.

[0706] Data Preprocessing

[0707] Server: Performs filtering to remove noise from the received video data. Removes digital noise and adjusts the contrast and brightness of the video.

[0708] Server: The filtered video data is converted to the optimal resolution and prepared for transmission to the analysis algorithm. Data is compressed and format converted.

[0709] Analysis for anomaly detection

[0710] Server: The filtered video data is fed into the AI ​​model and an object recognition algorithm is applied. For example, the AI ​​model identifies knives, suspicious liquids, etc.

[0711] Server: Then, behavioral recognition algorithms are used to detect certain anomalous behaviors (e.g., rapid movements or suspicious movements), and behavioral patterns are analyzed to identify behaviors that differ from normal behavior.

[0712] Analysis for Emotion Recognition

[0713] Server: Runs an emotion engine based on video data, analyzing the user's facial expressions, voice, and posture. For example, it can read emotions such as smile, anger, and sadness from camera footage.

[0714] Server: Emotion data determined by the emotion engine is supplied to the analysis means to improve the accuracy of anomaly detection. The emotion data adjusts the anomaly parameter, increasing the accuracy of abnormal behavior.

[0715] Identifying abnormal situations

[0716] Server: If an anomaly is detected based on the analysis results, a detailed report is generated, including the location, time, and type of anomaly.

[0717] Server: Converts reports into a user-friendly format and prepares them for notification. Automatically generates reports in text and visual formats.

[0718] Notification Processing

[0719] Server: Based on the generated report, it immediately notifies officials and police, specifically by launching a system to send emails, SMS, and in-app alerts.

[0720] Notification method: A notification is sent to the employee's device, an alert sound is heard to attract attention, and a pop-up notification is displayed on the PC screen of the employee in the monitoring room.

[0721] On-site response

[0722] User: Upon receiving the notification, staff and police rush to the scene to check the details of the abnormality. They head to the scene while checking the surveillance camera footage.

[0723] User: Eliminate the cause of the anomaly and take appropriate action to return the situation to safety. Staff will seize the dangerous object and take appropriate measures until the police arrive.

[0724] Video archiving and post-processing

[0725] Server: Stores video of detected anomalies for later analysis and as evidence. Stores video data in a secure storage and makes it accessible when needed.

[0726] Server: Uses the stored data to evaluate system performance and retrain the AI ​​model if necessary. Analyzes past anomaly detection cases and uses them as a dataset to improve the model's accuracy.

[0727] Specific examples

[0728] Specific example on a station platform

[0729] Terminal: Cameras installed on station platforms capture passenger activity in real time. Images of the platform during rush hour are stored in buffer memory.

[0730] Server: Receives video data and uses object and behavior recognition algorithms to detect people holding knives or making suspicious movements. The emotion engine then analyzes the person's emotions based on their facial expressions.

[0731] Server: Identifies a person holding a knife and with an angry expression as abnormal and generates a detailed report.

[0732] Notification method: If an abnormality is detected, an emergency notification will be sent to station staff and police via email and SMS. Staff will also receive a notification on their smartphones stating that "an angry person holding a knife is in the middle of the platform."

[0733] User: Staff rush to the scene and contact the police. Staff immediately head to the scene and handle the situation until the police arrive.

[0734] Example prompts to input to the generative AI model

[0735] "Analyze abnormal behavior and emotions based on video data collected by cameras installed on station platforms. Specifically, you need to detect people holding knives or behaving abnormally, and analyze their emotions as well."

[0736] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0737] Step 1:

[0738] Data collection

[0739] The terminal collects video in real time from surveillance cameras and sensors installed at various locations. Specifically, cameras installed on the station platform continuously capture video of the entire platform. (Input): Surveillance camera video data (Output): Acquired video data (stored in buffer memory)

[0740] The terminal stores the collected video data in buffer memory, making it available for immediate analysis.

[0741] Step 2:

[0742] Data Preprocessing

[0743] The server performs filtering on the video data received from the buffer memory to remove noise. Specifically, it removes digital noise from the video and adjusts the contrast and brightness. (Input): Video data in the buffer memory (Output): Video data after filtering

[0744] The server converts the filtered video data to the optimal resolution and prepares it for transmission to the analysis algorithm. Specific operations include data compression and format conversion. (Input): Filtered video data (Output): Resolution-converted video data

[0745] Step 3:

[0746] object recognition

[0747] The server inputs the resolution-converted video data into the AI ​​model and applies an object recognition algorithm. Specifically, the AI ​​model identifies knives, suspicious liquids, etc. (Input): Resolution-converted video data (Output): Object recognition results (data identifying knives, suspicious objects, etc.)

[0748] Step 4:

[0749] behavioral recognition

[0750] The server uses a behavior recognition algorithm based on the results of object recognition to detect specific abnormal behavior. Specifically, the AI ​​model analyzes behavior patterns and identifies behaviors that differ from normal behavior. (Input): Object recognition results (Output): Behavior recognition results (data identifying abnormal behavior)

[0751] Step 5:

[0752] emotion recognition

[0753] The server runs an emotion engine based on the behavioral recognition data, analyzing the user's facial expressions, voice, and posture. Specific behaviors include smiles, anger, sadness, and other emotions read from the camera footage. (Input): Behavior recognition results (Output): Emotion recognition results (data with identified emotions)

[0754] The server supplies the emotion data determined by the emotion engine to the analysis means to improve the accuracy of anomaly detection. (Input): Emotion recognition results (Output): Data with anomaly level parameters adjusted

[0755] Step 6:

[0756] Anomaly detection and report generation

[0757] If the server determines that an anomaly has occurred based on the analysis results, it generates a detailed report. Specifically, it creates a report that clearly indicates the location, time, and type of anomaly. (Input): Data with anomaly parameter adjustments (Output): Anomaly detection report

[0758] The server converts the generated report into an easy-to-read format and prepares it for notification. Specific operations include automatically generating reports in text and visual formats. (Input): Anomaly detection report (Output): Report that can be notified

[0759] Step 7:

[0760] Notification Processing

[0761] The server immediately notifies officials and police based on the contents of the notifiable report. Specific actions include launching a system to send emails, SMS, and in-app alerts. (Input): Notifiable report (Output): Sent notification (email, SMS, etc.)

[0762] The notification method is that a notification is sent to the employee's terminal and an alert sound is emitted to attract their attention. Specific actions include a pop-up notification being displayed on the PC screen in the monitoring room. (Input): Sent notification (Output): Notification displayed on the employee's terminal

[0763] Step 8:

[0764] On-site response

[0765] Upon receiving the notification, the staff or police rush to the scene and check the details of the abnormality. Specific actions include heading to the scene while checking the video from the surveillance camera. (Input): Notification displayed on staff's device (Output): Response begins at the scene

[0766] The user must take appropriate action to remove the cause of the anomaly and return the situation to safety. Specifically, the user must take appropriate action until the staff seizes the dangerous object and the police arrive. (Input): Start of on-site response (Output): Completion of anomaly response

[0767] Step 9:

[0768] Video archiving and post-processing

[0769] The server stores video of detected anomalies for later analysis and as evidence. Specific operations include storing the video data in a secure storage and making it accessible as needed. (Input): On-site response data (Output): Stored video of anomalies

[0770] The server uses the stored data to evaluate the system's performance and retrain the AI ​​model as necessary. Specifically, it analyzes past anomaly detection cases and uses them as a dataset to improve the model's accuracy. (Input): Stored anomaly video. (Output): Updated AI model.

[0771] (Application example 2)

[0772] 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."

[0773] Modern society requires rapid and appropriate responses to various abnormal situations that may occur in public places and educational institutions (for example, the appearance of a person armed with a knife or the abandonment of suspicious liquids). However, previous surveillance systems have focused on object recognition and behavioral recognition, making it difficult to combine emotion recognition for highly accurate anomaly detection. Furthermore, real-time notification capabilities have been limited, resulting in delayed responses when an abnormality occurs. This has sometimes led to insufficient safety in public places and educational institutions.

[0774] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0775] In this invention, the server includes a data acquisition means for collecting visual data for detecting anomalies, a data analysis means for detecting anomalies from the acquired visual data, a notification means for notifying anomalies detected by the data analysis means, an emotion recognition means for recognizing a user's emotion, and an integrated analysis means for determining a detection result by combining abnormal behavior and the user's emotion, thereby enabling highly accurate detection of anomalies and rapid response.

[0776] "Visual data" is digital data containing visual information that is collected by image capture devices such as cameras and sensors.

[0777] "Data acquisition means" is a general term for devices including cameras and various sensors installed to collect visual data.

[0778] "Data analysis means" includes software and hardware means for analyzing acquired visual data and detecting specific patterns or anomalies.

[0779] "Notification Methods" are the functions and devices used to notify personnel or law enforcement in real time of details of detected anomalies, including notifications in the form of email, SMS, in-app alerts, etc.

[0780] "Emotion recognition means" refers to algorithms and devices that analyze a user's facial expressions, voice, posture, etc., and identify their psychological state and emotions.

[0781] The "integrated analysis means" is a means for combining abnormal behavior and user emotion data to perform a comprehensive analysis and improve the accuracy of abnormality detection.

[0782] An "object recognition algorithm" is a computational procedure for detecting specific objects (e.g., knives, suspicious liquids) from video data.

[0783] A "behavior recognition algorithm" is a calculation procedure for analyzing the movements and actions of people in video data and identifying suspicious behavior.

[0784] The "storage means" is a device for storing video data of detected abnormalities for later analysis or use as evidence.

[0785] This invention is a system that quickly detects abnormalities and automates appropriate responses, improving safety in public places and educational institutions. This system is composed of a combination of data acquisition means, data analysis means, reporting means, emotion recognition means, and integrated analysis means.

[0786] Data Acquisition Method

[0787] The devices use cameras and sensors installed at each location to collect visual data, providing real-time footage of the entire public space or educational institution. For example, a camera installed on a train platform captures footage of the entire platform.

[0788] Data Analysis Methods

[0789] The server analyzes the acquired visual data. This analysis includes object recognition algorithms and behavior recognition algorithms. For example, it detects people holding knives or handling suspicious liquids. These object recognition and behavior recognition algorithms are implemented using frameworks such as OpenCV and Keras.

[0790] emotion recognition means

[0791] The server is equipped with emotion recognition means to analyze the user's facial expressions, voice, and posture, allowing it to identify a person's emotional state (e.g., anger, fear, sadness, etc.) The technologies used include deep learning models and supporting software such as TensorFlow and Keras models.

[0792] Integrated analysis method

[0793] The server combines abnormal behavior and user emotional data for comprehensive analysis, improving the accuracy of anomaly detection. This integrated analysis method further improves the accuracy of anomaly detection when, for example, the user is not only holding a knife but also has an angry expression.

[0794] Reporting means

[0795] The server provides a means of notification to notify personnel and police in real time of detected anomalies, including details of the anomalies, via email, SMS, in-app alerts, etc. Video data of the anomalies is stored in the cloud or local storage for later analysis or as evidence.

[0796] Specific examples

[0797] Cameras installed on station platforms capture passenger activity in real time. For example, the cameras capture a situation where there are many passengers on the platform during rush hour. The server receives the video data and first uses object recognition and behavior recognition algorithms to detect people holding knives or people behaving suspiciously. The server then analyzes the person's emotions from facial expressions captured by emotion recognition means. If a person holding a knife or with an angry expression is judged to be abnormal, an emergency notification is immediately sent to staff and the police. Staff receive the notification and rush to the scene.

[0798] Prompt Sentence Examples

[0799] It analyzes camera footage in real time to detect suspicious individuals and movements. It also uses facial recognition technology to analyze emotions and immediately notify staff and police if an abnormality is detected. Please list the hardware and software required for this system and explain how to implement it specifically.

[0800] This system is expected to enable rapid and highly accurate detection of abnormalities in public places and educational institutions, significantly improving safety.

[0801] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0802] Step 1:

[0803] The device uses cameras and sensors installed at each location to collect visual data. The cameras and sensors capture video data in real time and temporarily store the data in local storage. The input is the visual data from the camera, and the output is the temporarily stored video data.

[0804] Step 2:

[0805] The server preprocesses the video data received from the terminal. First, it filters the data to remove noise, and then converts it to the required resolution. This process prepares the data for the data analysis means to operate efficiently. The input is the temporarily stored video data, and the output is the filtered and resolution-converted video data.

[0806] Step 3:

[0807] The server analyzes the filtered video data. In this step, it applies object and behavior recognition algorithms to detect abnormal behaviors and objects. For example, it identifies people holding knives or suspicious liquids. The input is the filtered and resolution-converted video data, and the output is the detection result data indicating anomalies.

[0808] Step 4:

[0809] The server analyzes the user's emotions based on the detection result data. Using emotion recognition means, it analyzes the user's facial expressions, voice, posture, etc., and identifies emotions such as anger, sadness, and fear. The input is the detection result data and visual data regarding abnormal behavior and abnormal objects, and the output is the emotion recognition result data.

[0810] Step 5:

[0811] The server integrates the abnormal behavior and emotion recognition results to make a final anomaly determination. Specifically, if a person holding a knife also displays anger, it is judged to be highly abnormal. The input is abnormal behavior data and emotion recognition result data, and the output is the integrated anomaly detection result.

[0812] Step 6:

[0813] The server activates the notification mechanism based on the final anomaly detection results. Emergency notifications are sent to personnel and police in real time, and detailed reports are generated. Notification formats include email, SMS, and in-app alerts, and messages containing anomaly identification information are sent. The input is the consolidated anomaly detection results, and the output is emergency notifications and detailed reports.

[0814] Step 7:

[0815] The user receives notifications from the reporting means and responds promptly to the scene. Specifically, when a staff member receives a notification, he or she immediately heads to the scene to confirm and respond to the abnormality. The input is the emergency notification and detailed report, and the output is the actual abnormality response action.

[0816] Step 8:

[0817] The server stores video data when an anomaly is detected. The stored data is used for later analysis and as evidence. The input is the video data when an anomaly occurs, and the output is the data stored in the cloud or local storage.

[0818] This series of processing steps enables the system to detect anomalies with high accuracy and respond quickly, significantly improving safety in public places and educational institutions.

[0819] 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.

[0820] 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.

[0821] 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.

[0822] [Third embodiment]

[0823] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0824] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0825] 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).

[0826] 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.

[0827] 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.

[0828] 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).

[0829] 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.

[0830] 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.

[0831] 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.

[0832] 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.

[0833] 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.

[0834] 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."

[0835] This invention is a system that quickly detects abnormalities in public places and educational institutions and automates appropriate responses. This system combines video capture, analysis, and notification means to significantly improve safety at the site.

[0836] Program processing

[0837] 1. Data Collection

[0838] Terminal: Collects real-time video from surveillance cameras and sensors installed in various locations. For example, a camera installed on a station platform captures video of the entire platform.

[0839] Terminal: Temporarily stores collected video data and makes it available for immediate analysis. For example, video captured by a camera is temporarily stored in local storage.

[0840] 2. Data Preprocessing

[0841] Server: Filters the received video data to remove noise, converting it into a clear, easy-to-analyze image.

[0842] Server: Converts the filtered video data to the optimal resolution and prepares it for sending to the analysis means.

[0843] 3. Analysis for Anomaly Detection

[0844] Server: The filtered video data is fed into an AI model, where object and behavior recognition algorithms are applied, for example to detect people holding knives or handling suspicious liquids.

[0845] Server: Identifies specific abnormal behavior, such as detecting someone walking quickly or moving differently from the normal crowd.

[0846] 4. Identifying abnormal situations

[0847] Server: If an anomaly is detected based on the analysis results, a detailed report is generated, including the location, time, and type of anomaly detected.

[0848] Server: Transforms the report into a readable format and prepares it for notification.

[0849] 5. Notification Processing

[0850] Server: Based on the generated report, it immediately notifies officials and police via email, SMS, in-app alerts, etc.

[0851] Notification method: A notification is sent to the employee's device and an alert sound is generated to attract attention. For example, a notification is displayed on the computer of an employee working in the monitoring room.

[0852] 6. On-site response

[0853] User: Upon receiving the notification, staff or police rush to the scene and check the details of the abnormality. For example, staff rush to the scene while checking the surveillance camera footage.

[0854] User: Eliminate the cause of the anomaly and take appropriate action to return the situation to safety.

[0855] 7. Video storage and post-processing

[0856] Server: Stores video footage of detected anomalies for later analysis or as evidence. For example, it stores video footage of before and after the anomaly occurred in a specific format.

[0857] Server: Analyzes the stored video data and uses it to evaluate and improve system performance.

[0858] Specific examples

[0859] Example on a station platform:

[0860] Terminal: Cameras installed on station platforms capture passenger activity in real time. For example, cameras can capture the situation of large numbers of passengers on the platform during rush hour.

[0861] Server: Receives video data and analyzes abnormal behavior, such as detecting someone pulling out a knife in a crowd.

[0862] Notification method: If an abnormality is detected, an emergency notification will be sent to station staff and the police via email or SMS. For example, a staff member may receive a notification on their smartphone stating that "a person with a knife is in the center of the platform."

[0863] User: Staff rush to the scene and contact the police. For example, staff immediately head to the scene and handle the situation until the police arrive.

[0864] Thus, the present invention is a system that provides an embodiment for quickly detecting abnormalities in public places and educational institutions and taking appropriate measures.

[0865] The processing flow will be explained below.

[0866] Step 1:

[0867] Terminal: Collects real-time video from surveillance cameras and sensors installed in various locations. For example, a camera installed on a station platform captures video of the entire platform.

[0868] Step 2:

[0869] Terminal: Collected video data is temporarily stored in local storage, making it available for immediate analysis. For example, video captured by a camera is temporarily stored in a buffer.

[0870] Step 3:

[0871] Server: Performs filtering to remove noise from the received video data, for example, removing non-essential parts of the video data and converting it into a clear, easy-to-analyze format.

[0872] Step 4:

[0873] Server: Converts filtered video data to the optimal resolution and prepares it for sending to the analysis algorithm. Converts data into the format required by the analysis method.

[0874] Step 5:

[0875] Server: The filtered video data is fed into an AI model, where object and behavior recognition algorithms are applied, for example to detect people holding knives or handling suspicious liquids.

[0876] Step 6:

[0877] Server: Identifies specific abnormal behavior, such as people walking quickly or moving differently from the general crowd, thereby identifying suspicious individuals in real time.

[0878] Step 7:

[0879] Server: If an anomaly is detected based on the analysis results, a detailed report is generated, including the location, time, and type of anomaly detected.

[0880] Step 8:

[0881] Server: Transforms reports into a format that is easy to read and prepares them for notification, e.g., creating text and visual reports.

[0882] Step 9:

[0883] Server: Based on the generated report, it immediately notifies officials and police via email, SMS, in-app alerts, etc.

[0884] Step 10:

[0885] Terminal: A notification is sent to the employee's terminal and an alert sound is generated to attract attention. For example, a notification is displayed on the computer of an employee working in the monitoring room.

[0886] Step 11:

[0887] User: Upon receiving the notification, staff or police rush to the scene and check the details of the abnormality. For example, staff rush to the scene while checking the surveillance camera footage.

[0888] Step 12:

[0889] User: Eliminate the cause of the anomaly and take appropriate action to return the situation to safety. For example, staff will seize the dangerous material and coordinate with the police.

[0890] Step 13:

[0891] Server: Stores video footage of detected anomalies for later analysis or as evidence. For example, it stores video footage of before and after the anomaly occurred in a specific format.

[0892] Step 14:

[0893] Server: Analyzes the stored video data to evaluate and improve system performance, for example by retraining the AI ​​model based on past cases.

[0894] Example 1

[0895] 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."

[0896] Public places and educational institutions require the ability to quickly and accurately detect anomalies and automate appropriate responses. However, conventional systems lack the accuracy of anomaly detection and notification speed required, making manual monitoring and response necessary. Furthermore, the process of storing detected anomalies and post-analysis was often inefficient. A system that solves these problems and significantly improves safety is needed.

[0897] 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.

[0898] In this invention, the server includes a data acquisition means for collecting video data for detecting anomalies, a data preprocessing means for removing noise from the acquired video data and optimizing resolution, an analysis means for detecting anomalies from the video data processed by the data preprocessing means, a notification means for notifying anomalies detected by the analysis means, and a video storage means for storing video of the detected anomalies for use as later analysis or evidence. This enables rapid and accurate detection of anomalies, real-time notification and storage, and further subsequent analysis.

[0899] "Data acquisition means" refers to the devices and technologies that collect the video data necessary to detect abnormalities.

[0900] "Data pre-processing means" refers to technology or devices that perform processing to remove noise from acquired video data and optimize resolution.

[0901] "Analysis means" refers to a device that includes algorithms and techniques for detecting anomalies based on pre-processed video data.

[0902] "Notification means" refers to the technology or device used to notify relevant parties of abnormalities detected by the analysis means.

[0903] "Video storage means" refers to technology or equipment for storing video of detected abnormalities and using it for later analysis or as evidence.

[0904] "Object recognition algorithm" refers to the computational processes and techniques used to detect and identify objects based on video data.

[0905] "Behavior recognition algorithm" refers to the computational processes and technologies used to detect and identify behaviors based on video data and determine abnormal behavior.

[0906] "Emergency notification function" refers to technology and systems that notify relevant parties such as staff and police in real time of details of detected abnormalities.

[0907] The present invention relates to a system for quickly detecting abnormalities in public places and educational institutions and automating appropriate responses. The system includes a data acquisition unit, a data preprocessing unit, an analysis unit, a notification unit, and a video storage unit.

[0908] The system is programmed as follows:

[0909] First, surveillance cameras and other data acquisition methods installed at various locations collect video data in real time. For example, cameras installed on train platforms capture passenger movements.

[0910] The device then temporarily stores the collected video data and makes it available for immediate analysis using local storage.

[0911] The server then performs a filtering process to remove noise from the received video data, making the video clearer and easier to analyze. The filtered video data is then converted to an optimal resolution and prepared for transmission to the analysis means.

[0912] The analysis method inputs the filtered video data into an AI model and applies object and behavior recognition algorithms to detect, for example, people holding knives or handling suspicious liquids, as well as identify specific abnormal behaviors, such as people walking quickly or moving differently from the general crowd.

[0913] If the analysis results in an anomaly, the server generates a detailed report that includes the location, time, and type of anomaly detected, and converts the report into an easy-to-read format for notification.

[0914] The notification method will immediately notify employees and police based on the contents of the generated report. Notifications will be sent via email, SMS, in-app alerts, etc. For example, an emergency notification such as "There is a person with a knife on the station platform" will be sent to an employee's smartphone.

[0915] The users who receive the notification, i.e., officials or police, rush to the scene and check the details of the abnormality. Officials rush to the scene while checking the surveillance camera footage and take appropriate action to eliminate the cause of the abnormality.

[0916] Finally, the server stores the video of the detected anomaly for later analysis and as evidence. The stored video data can be analyzed to evaluate and improve the system's performance.

[0917] Prompt Sentence Examples

[0918] Below are some example prompts to input to the generative AI model:

[0919] "Please explain the most effective method for detecting abnormal behavior in public places."

[0920] "What are the key elements of a system that quickly detects anomalies and automatically notifies you?"

[0921] Thus, the present invention is a system that provides an embodiment for quickly detecting abnormalities in public places and educational institutions and taking appropriate measures.

[0922] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0923] Step 1: Data collection

[0924] Terminal: Surveillance cameras installed in various locations collect video data in real time. For example, cameras installed on train platforms capture passenger movements.

[0925] Input: Real-time video data from public places and educational institutions.

[0926] Output: Real-time video data.

[0927] Step 2: Temporarily save the video data

[0928] Terminal: Collected video data is temporarily stored in local storage. For example, video data captured by a camera is temporarily stored on a local drive so that it can be analyzed immediately.

[0929] Input: Real-time video data.

[0930] Output: Stored temporary video data.

[0931] Step 3: Filtering

[0932] Server: Filters the received video data to remove noise. By removing unnecessary noise and blur, the video becomes clear and easier to analyze.

[0933] Input: Stored temporary video data.

[0934] Output: Filtered video data.

[0935] Step 4: Optimize the resolution

[0936] Server: Converts the filtered video data to the appropriate resolution and prepares it for analysis. For example, compresses high-resolution video slightly to maintain analysis accuracy and speed.

[0937] Input: Filtered video data.

[0938] Output: Optimized resolution video data.

[0939] Step 5: Input to the AI ​​model

[0940] Server: The optimized resolution video data is input into the AI ​​model and an object recognition algorithm is applied to detect people and objects in the video data.

[0941] Input: Video data with optimized resolution.

[0942] Output: Object and person detection results.

[0943] Step 6: Applying the activity recognition algorithm

[0944] Server: Uses the results of the object recognition algorithm to apply the behavior recognition algorithm, for example, to detect people whose movements deviate from their normal behavioral patterns.

[0945] Input: Object and person detection results.

[0946] Output: Abnormal behavior detection results.

[0947] Step 7: Generate anomaly detection reports

[0948] Server: If the analysis results in an anomaly, a detailed report is generated, including the location, time, and specific details of the anomaly.

[0949] Input: Anomalous behavior detection results.

[0950] Output: Anomaly detection report.

[0951] Step 8: Convert the report to another format

[0952] Server: Converts the generated reports into a format that is easy to read, for example, creating reports in PDF or HTML format.

[0953] Input: Anomaly detection report.

[0954] Output: Formatted report.

[0955] Step 9: Call Emergency Services

[0956] Server: Based on the generated reports, it provides instant notifications to officials and police via email, SMS or in-app alerts.

[0957] Input: The formatted report.

[0958] Output: Emergency call notification.

[0959] Step 10: Displaying notifications

[0960] Notification method: An emergency notification will be displayed on the employee's terminal and an alert sound will be emitted to attract attention. For example, the notification will be displayed immediately on the computer of an employee on duty in the monitoring room.

[0961] Input: Emergency call notification.

[0962] Output: Notifications that appear on the device.

[0963] Step 11: Rush to the scene

[0964] User: Upon receiving the notification, staff and police rush to the scene to check the details of the abnormality. Staff head to the scene while checking the surveillance camera footage.

[0965] Input: Notifications that appear on your device.

[0966] Output: Check the details of the error.

[0967] Step 12: Dealing with abnormalities

[0968] User: Eliminate the cause of the anomaly and take appropriate action to return the situation to safety. For example, staff may apprehend the anomaly.

[0969] Input: Check the details of the abnormality.

[0970] Output: Safe state restored.

[0971] Step 13: Save the abnormal video

[0972] Server: Stores video footage of detected anomalies for later analysis and as evidence. Stores footage of several minutes before and after the anomaly occurs.

[0973] Input: Detailed confirmation of anomaly and video data.

[0974] Output: Archived footage of the anomaly.

[0975] Step 14: Performance evaluation and system improvement

[0976] Server: Analyzes the stored video data to evaluate and improve the system's performance. For example, it evaluates detection accuracy and notification speed, and improves the AI ​​model.

[0977] Input: Archived anomalous footage.

[0978] Output: System improvement measures and evaluation results.

[0979] (Application example 1)

[0980] 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."

[0981] In modern society, there is a demand for immediate detection and rapid response of abnormal behavior in public places and educational institutions. However, current systems can be slow to detect and notify abnormalities, or on-site staff may not be able to respond immediately. Furthermore, there is a lack of means to immediately communicate detailed information about abnormalities to staff visually and audibly, making it difficult to respond to abnormalities efficiently. The objective of the present invention is to solve these problems.

[0982] 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.

[0983] In this invention, the server includes a video acquisition means for collecting camera data for detecting abnormalities, an analysis means for detecting abnormalities from the acquired camera data, a notification means for notifying the user of the abnormality detected by the analysis means, and a wearable display means for immediately conveying the notification information from the notification means to the user's visual and auditory senses. This makes it possible to quickly detect abnormal behavior, notify staff in real time, and respond immediately on site.

[0984] "Video acquisition means" refers to equipment and software that collects camera data for detecting abnormalities.

[0985] "Analysis means" refers to algorithms and software for detecting anomalies in acquired camera data.

[0986] The "notification means" refers to a system and software for notifying staff and other relevant parties of abnormalities detected by the analysis means.

[0987] The "wearable display means" is a device for instantly conveying the notification information of the notification means to the user's visual and auditory senses.

[0988] An "object recognition algorithm" is a program for identifying specific objects (e.g., knives, suspicious liquids) from video data.

[0989] A "behavior recognition algorithm" is a program that analyzes people's movements from video data and detects abnormal behavior (e.g., suspicious movements).

[0990] The "emergency notification function" is a function that notifies staff and police in real time of detected abnormalities.

[0991] A "server" is a main computing device that performs a series of processes such as data collection, analysis, and notification.

[0992] "User" refers to the staff and other relevant parties who receive and respond to the notified abnormality information.

[0993] The system for realizing this invention is configured using the following hardware and software: The hardware used is a surveillance camera, a cloud server, and smart glasses, while the software used is an AI model for video analysis (e.g., YOLO, OpenPose), a library for video noise reduction and filtering (e.g., OpenCV), and a notification service API (e.g., Firebase Cloud Messaging, Twilio).

[0994] The server collects video data from surveillance cameras in real time, which is first pre-processed with a program to remove noise and perform filtering, improving the quality of the video data and making it easier to analyze.

[0995] The server then inputs the preprocessed video data into an AI model and applies object and behavior recognition algorithms, such as YOLO to detect knives and suspicious liquids and OpenPose to identify abnormal behaviors, enabling the detection of specific abnormal behaviors and objects in real time.

[0996] If the server detects an anomaly based on these analysis results, it generates a detailed report containing information such as the location, time, and type of anomaly. After the report is generated, an emergency notification is sent immediately to staff and relevant parties via the notification service API. Notifications are sent in multiple ways, including email, SMS, and in-app alerts.

[0997] The user, a staff member, receives notifications through the smart glasses. The wearable display immediately conveys the notification information visually and audibly, allowing the staff member to quickly grasp the situation on-site. The staff member can check the details of any abnormalities in real time and take prompt action.

[0998] Examples:

[0999] For example, if a surveillance camera installed on a station platform detects a person holding a knife in the crowd, the server will filter the footage and identify the person as holding a knife using the YOLO model. If an anomaly is detected, the server will generate a detailed report and send a notification to the employee's smart glasses via the notification service API, stating, "There is a person holding a knife in the center of the platform." The employee will receive the notification and confirm the visual and audio alert, allowing them to respond to the scene quickly.

[1000] Example prompt sentence:

[1001] "Analyze the footage from the center of the platform to detect if there is anyone holding a knife."

[1002] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1003] Step 1:

[1004] The device transmits real-time video data captured by the camera to the cloud server. The input is video data from the surveillance camera, and the output is temporarily stored video data. This process involves acquiring video data and immediately transferring it to the server.

[1005] Step 2:

[1006] The server performs filtering to remove noise from the received video data. The input is the temporarily stored video data, and the output is the noise-removed video data. This process uses the OpenCV library to process the data to improve the video quality.

[1007] Step 3:

[1008] The server converts the filtered video data to the optimal resolution and prepares it for transmission to the analysis means. The input is the noise-removed video data, and the output is video data converted to a resolution suitable for analysis. This process involves data calculations using high-performance image processing algorithms.

[1009] Step 4:

[1010] The server then inputs the converted video data into an AI model and applies object and action recognition algorithms. The input is video data suitable for analysis, and the output is detected anomaly information. This process uses YOLO and OpenPose for object and action recognition and data calculation.

[1011] Step 5:

[1012] The server determines anomalies based on the analysis results of the AI ​​model and generates a detailed report. The input is the detected anomaly information, and the output is a report summarizing the details of the anomaly. This process generates a document summarizing the type of anomaly, the location where it occurred, the time, etc.

[1013] Step 6:

[1014] Based on the generated report, the server sends emergency notifications to staff and related parties via the notification service API. The input is a detailed report, and the output is the sent notification. This process uses services such as Firebase Cloud Messaging and Twilio to send notifications immediately.

[1015] Step 7:

[1016] The user receives notifications through smart glasses and acknowledges visual and audible alerts. The input is the notification sent, and the output is the user's acknowledgement and immediate response. This process uses the wearable device to display information and provide alerts.

[1017] Step 8:

[1018] After checking the contents of the notification, the user quickly heads to the site, checks the details of the abnormality, and takes action. The input is the content of the notification and the situation at the site, and the output is the appropriate action to eliminate the cause of the abnormality. This process allows for a quick response and problem resolution at the site.

[1019] 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.

[1020] This invention is a system that quickly detects abnormalities in public places and educational institutions and automates appropriate responses. This system combines video acquisition means, analysis means, notification means, and an emotion engine that recognizes user emotions, significantly improving safety.

[1021] Program processing

[1022] 1. Data Collection

[1023] Terminal: Collects real-time video from surveillance cameras and sensors installed in various locations. For example, a camera installed on a station platform captures video of the entire platform.

[1024] Terminal: Collected video data is temporarily stored in local storage, making it available for immediate analysis. For example, video captured by a camera is temporarily stored in a buffer.

[1025] 2. Data Preprocessing

[1026] Server: Performs filtering to remove noise from the received video data, for example, removing non-essential parts of the video data and converting it into a clear, easy-to-analyze format.

[1027] Server: Converts filtered video data to the optimal resolution and prepares it for sending to the analysis algorithm. Converts data into the format required by the analysis method.

[1028] 3. Analysis for Anomaly Detection

[1029] Server: The filtered video data is fed into an AI model, where object and behavior recognition algorithms are applied, for example to detect people holding knives or handling suspicious liquids.

[1030] Server: Identifies specific abnormal behavior, such as people walking quickly or moving differently from the general crowd, thereby identifying suspicious individuals in real time.

[1031] 4. Analysis for Emotion Recognition

[1032] Server: Runs an emotion engine based on video data and analyzes the user's facial expressions, voice, posture, etc. For example, it analyzes emotions such as smile, anger, and sadness from camera footage.

[1033] Server: Emotion data determined by the emotion engine is supplied to the analysis means and used as data to improve the accuracy of anomaly detection. For example, if a person with an angry expression is holding a knife, the degree of anomaly is set high.

[1034] 5. Identifying abnormal situations

[1035] Server: If an anomaly is detected based on the analysis results, a detailed report is generated, including the location, time, and type of anomaly detected.

[1036] Server: Transforms reports into a format that is easy to read and prepares them for notification, e.g., creating text and visual reports.

[1037] 6. Notification Processing

[1038] Server: Based on the generated report, it immediately notifies officials and police via email, SMS, in-app alerts, etc.

[1039] Notification method: A notification is sent to the employee's device and an alert sound is generated to attract attention. For example, a notification is displayed on the computer of an employee working in the monitoring room.

[1040] 7. On-site response

[1041] User: Upon receiving the notification, staff or police rush to the scene and check the details of the abnormality. For example, staff rush to the scene while checking the surveillance camera footage.

[1042] User: Eliminate the cause of the anomaly and take appropriate action to return the situation to safety. For example, staff will seize the dangerous material and coordinate with the police.

[1043] 8. Video storage and post-processing

[1044] Server: Stores video footage of detected anomalies for later analysis or as evidence. For example, it stores video footage of before and after the anomaly occurred in a specific format.

[1045] Server: Analyzes the stored video data to evaluate and improve system performance, for example by retraining the AI ​​model based on past cases.

[1046] Specific examples

[1047] Example on a station platform:

[1048] Terminal: Cameras installed on station platforms capture passenger activity in real time. For example, cameras can capture the situation of large numbers of passengers on the platform during rush hour.

[1049] Server: After receiving the video data, the server first uses object recognition and behavior recognition algorithms to detect people holding a knife or making suspicious movements.Then, the server analyzes the person's emotions based on facial expressions captured by the emotion engine.

[1050] Server: A person holding a knife and with an angry expression is judged to be abnormal.

[1051] Notification method: If an abnormality is detected, an emergency notification will be sent to station staff and the police via email or SMS. For example, a staff member may receive a notification on their smartphone saying, "There is an angry person holding a knife in the middle of the platform."

[1052] User: Staff rush to the scene and contact the police. For example, staff immediately head to the scene and handle the situation until the police arrive.

[1053] In this way, by combining emotion engines, the present invention is a system that provides an embodiment for detecting anomalies with higher accuracy and taking prompt and appropriate action.

[1054] The processing flow will be explained below.

[1055] Step 1:

[1056] Terminal: Collects real-time video from surveillance cameras and sensors installed in various locations. For example, a camera installed on a station platform captures video of the entire platform.

[1057] Step 2:

[1058] Terminal: Collected video data is temporarily stored in local storage, making it available for immediate analysis. For example, video captured by a camera is temporarily stored in a buffer.

[1059] Step 3:

[1060] Server: Performs filtering to remove noise from the received video data, for example, removing non-essential parts of the video data and converting it into a clear, easy-to-analyze format.

[1061] Step 4:

[1062] Server: Converts filtered video data to the optimal resolution and prepares it for sending to the analysis algorithm. Converts data into the format required by the analysis method.

[1063] Step 5:

[1064] Server: The filtered video data is fed into an AI model, where object and behavior recognition algorithms are applied, for example to detect people holding knives or handling suspicious liquids.

[1065] Step 6:

[1066] Server: Identifies specific abnormal behavior, such as people walking quickly or moving differently from the general crowd, thereby identifying suspicious individuals in real time.

[1067] Step 7:

[1068] Server: Runs an emotion engine based on video data and analyzes the user's facial expressions, voice, posture, etc. For example, it analyzes emotions such as smile, anger, and sadness from camera footage.

[1069] Step 8:

[1070] Server: Emotion data determined by the emotion engine is supplied to the analysis means and used as data to improve the accuracy of anomaly detection. For example, if a person with an angry expression is holding a knife, the degree of anomaly is set high.

[1071] Step 9:

[1072] Server: If an anomaly is detected based on the analysis results, a detailed report is generated, including the location, time, and type of anomaly detected.

[1073] Step 10:

[1074] Server: Transforms reports into a format that is easy to read and prepares them for notification, e.g., creating text and visual reports.

[1075] Step 11:

[1076] Server: Based on the generated report, it immediately notifies officials and police via email, SMS, in-app alerts, etc.

[1077] Step 12:

[1078] Terminal: A notification is sent to the employee's terminal and an alert sound is generated to attract attention. For example, a notification is displayed on the computer of an employee working in the monitoring room.

[1079] Step 13:

[1080] User: Upon receiving the notification, staff or police rush to the scene and check the details of the abnormality. For example, staff rush to the scene while checking the surveillance camera footage.

[1081] Step 14:

[1082] User: Eliminate the cause of the anomaly and take appropriate action to return the situation to safety. For example, staff will seize the dangerous material and coordinate with the police.

[1083] Step 15:

[1084] Server: Stores video footage of detected anomalies for later analysis or as evidence. For example, it stores video footage of before and after the anomaly occurred in a specific format.

[1085] Step 16:

[1086] Server: Analyzes the stored video data to evaluate and improve system performance, for example by retraining the AI ​​model based on past cases.

[1087] Example 2

[1088] 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."

[1089] Public places and educational institutions are required to quickly detect suspicious individuals and abnormal behavior and respond appropriately. However, conventional systems have low anomaly detection accuracy and frequently generate false alarms. Furthermore, they are unable to respond quickly in emergencies, making it difficult to prevent serious damage. Therefore, there is a need to develop a new system that improves the accuracy of anomaly detection and enables quick and appropriate responses.

[1090] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes an acquisition means for collecting data, a preprocessing means for filtering the acquired data and converting its resolution, and an emotion recognition means for recognizing the user's emotion based on the preprocessed data. This makes it possible to analyze abnormal behavior and the user's emotion with high accuracy and to take prompt and appropriate action.

[1091] An "acquisition means" is a device or group of devices that collects data to detect abnormalities.

[1092] The "analysis means" refers to a processing device or algorithm for detecting anomalies based on the acquired data.

[1093] The "notification means" refers to a device or software that notifies relevant parties such as staff and police of abnormalities detected by the analysis means.

[1094] The "pre-processing means" refers to a processing device or algorithm that filters the acquired image data and converts it to an optimal resolution.

[1095] "Emotion recognition means" refers to an algorithm or device for analyzing the user's emotions based on acquired video data.

[1096] An "object recognition algorithm" is a program or processing method for analyzing video data and identifying specific objects (e.g., knives, suspicious liquids).

[1097] A "behavior recognition algorithm" is a program or processing method for analyzing video data and detecting specific behaviors (e.g., suspicious movements).

[1098] The "emergency notification function" is a function that sends real-time notifications to staff and police when an abnormality is detected.

[1099] This invention is a system that quickly detects abnormalities in public places and educational institutions and automates appropriate responses. This system combines video acquisition means, analysis means, notification means, and an emotion engine that recognizes user emotions, significantly improving safety.

[1100] The overall configuration of the system includes an "acquisition means" that collects data to detect abnormalities, a "preprocessing means" that filters and converts the resolution of the acquired data, an "analysis means" that analyzes the preprocessed data, a "notification means" that notifies the user of the analysis results, and an "emotion recognition means" that recognizes the user's emotions.

[1101] Explanation of program processing

[1102] Below is a detailed description of the process at each stage of the system.

[1103] Data collection

[1104] Terminal: Surveillance cameras and sensors installed in various locations collect video in real time. For example, a camera installed on a station platform captures video of the entire platform. These videos are temporarily stored in local storage.

[1105] Terminal: Collected video data is stored in buffer memory so that it can be analyzed immediately.

[1106] Data Preprocessing

[1107] Server: Performs filtering to remove noise from the received video data. Removes digital noise and adjusts the contrast and brightness of the video.

[1108] Server: The filtered video data is converted to the optimal resolution and prepared for transmission to the analysis algorithm. Data is compressed and format converted.

[1109] Analysis for anomaly detection

[1110] Server: The filtered video data is fed into the AI ​​model and an object recognition algorithm is applied. For example, the AI ​​model identifies knives, suspicious liquids, etc.

[1111] Server: Then, behavioral recognition algorithms are used to detect certain anomalous behaviors (e.g., rapid movements or suspicious movements), and behavioral patterns are analyzed to identify behaviors that differ from normal behavior.

[1112] Analysis for Emotion Recognition

[1113] Server: Runs an emotion engine based on video data, analyzing the user's facial expressions, voice, and posture. For example, it can read emotions such as smile, anger, and sadness from camera footage.

[1114] Server: Emotion data determined by the emotion engine is supplied to the analysis means to improve the accuracy of anomaly detection. The emotion data adjusts the anomaly parameter, increasing the accuracy of abnormal behavior.

[1115] Identifying abnormal situations

[1116] Server: If an anomaly is detected based on the analysis results, a detailed report is generated, including the location, time, and type of anomaly.

[1117] Server: Converts reports into a user-friendly format and prepares them for notification. Automatically generates reports in text and visual formats.

[1118] Notification Processing

[1119] Server: Based on the generated report, it immediately notifies officials and police, specifically by launching a system to send emails, SMS, and in-app alerts.

[1120] Notification method: A notification is sent to the employee's device, an alert sound is heard to attract attention, and a pop-up notification is displayed on the PC screen of the employee in the monitoring room.

[1121] On-site response

[1122] User: Upon receiving the notification, staff and police rush to the scene to check the details of the abnormality. They head to the scene while checking the surveillance camera footage.

[1123] User: Eliminate the cause of the anomaly and take appropriate action to return the situation to safety. Staff will seize the dangerous object and take appropriate measures until the police arrive.

[1124] Video archiving and post-processing

[1125] Server: Stores video of detected anomalies for later analysis and as evidence. Stores video data in a secure storage and makes it accessible when needed.

[1126] Server: Uses the stored data to evaluate system performance and retrain the AI ​​model if necessary. Analyzes past anomaly detection cases and uses them as a dataset to improve the model's accuracy.

[1127] Specific examples

[1128] Specific example on a station platform

[1129] Terminal: Cameras installed on station platforms capture passenger activity in real time. Images of the platform during rush hour are stored in buffer memory.

[1130] Server: Receives video data and uses object and behavior recognition algorithms to detect people holding knives or making suspicious movements. The emotion engine then analyzes the person's emotions based on their facial expressions.

[1131] Server: Identifies a person holding a knife and with an angry expression as abnormal and generates a detailed report.

[1132] Notification method: If an abnormality is detected, an emergency notification will be sent to station staff and police via email and SMS. Staff will also receive a notification on their smartphones stating that "an angry person holding a knife is in the middle of the platform."

[1133] User: Staff rush to the scene and contact the police. Staff immediately head to the scene and handle the situation until the police arrive.

[1134] Example prompts to input to the generative AI model

[1135] "Analyze abnormal behavior and emotions based on video data collected by cameras installed on station platforms. Specifically, you need to detect people holding knives or behaving abnormally, and analyze their emotions as well."

[1136] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1137] Step 1:

[1138] Data collection

[1139] The terminal collects video in real time from surveillance cameras and sensors installed at various locations. Specifically, cameras installed on the station platform continuously capture video of the entire platform. (Input): Surveillance camera video data (Output): Acquired video data (stored in buffer memory)

[1140] The terminal stores the collected video data in buffer memory, making it available for immediate analysis.

[1141] Step 2:

[1142] Data Preprocessing

[1143] The server performs filtering on the video data received from the buffer memory to remove noise. Specifically, it removes digital noise from the video and adjusts the contrast and brightness. (Input): Video data in the buffer memory (Output): Video data after filtering

[1144] The server converts the filtered video data to the optimal resolution and prepares it for transmission to the analysis algorithm. Specific operations include data compression and format conversion. (Input): Filtered video data (Output): Resolution-converted video data

[1145] Step 3:

[1146] object recognition

[1147] The server inputs the resolution-converted video data into the AI ​​model and applies an object recognition algorithm. Specifically, the AI ​​model identifies knives, suspicious liquids, etc. (Input): Resolution-converted video data (Output): Object recognition results (data identifying knives, suspicious objects, etc.)

[1148] Step 4:

[1149] behavioral recognition

[1150] The server uses a behavior recognition algorithm based on the results of object recognition to detect specific abnormal behavior. Specifically, the AI ​​model analyzes behavior patterns and identifies behaviors that differ from normal behavior. (Input): Object recognition results (Output): Behavior recognition results (data identifying abnormal behavior)

[1151] Step 5:

[1152] emotion recognition

[1153] The server runs an emotion engine based on the behavioral recognition data, analyzing the user's facial expressions, voice, and posture. Specific behaviors include smiles, anger, sadness, and other emotions read from the camera footage. (Input): Behavior recognition results (Output): Emotion recognition results (data with identified emotions)

[1154] The server supplies the emotion data determined by the emotion engine to the analysis means to improve the accuracy of anomaly detection. (Input): Emotion recognition results (Output): Data with anomaly level parameters adjusted

[1155] Step 6:

[1156] Anomaly detection and report generation

[1157] If the server determines that an anomaly has occurred based on the analysis results, it generates a detailed report. Specifically, it creates a report that clearly indicates the location, time, and type of anomaly. (Input): Data with anomaly parameter adjustments (Output): Anomaly detection report

[1158] The server converts the generated report into an easy-to-read format and prepares it for notification. Specific operations include automatically generating reports in text and visual formats. (Input): Anomaly detection report (Output): Report that can be notified

[1159] Step 7:

[1160] Notification Processing

[1161] The server immediately notifies officials and police based on the contents of the notifiable report. Specific actions include launching a system to send emails, SMS, and in-app alerts. (Input): Notifiable report (Output): Sent notification (email, SMS, etc.)

[1162] The notification method is that a notification is sent to the employee's terminal and an alert sound is emitted to attract their attention. Specific actions include a pop-up notification being displayed on the PC screen in the monitoring room. (Input): Sent notification (Output): Notification displayed on the employee's terminal

[1163] Step 8:

[1164] On-site response

[1165] Upon receiving the notification, the staff or police rush to the scene and check the details of the abnormality. Specific actions include heading to the scene while checking the video from the surveillance camera. (Input): Notification displayed on staff's device (Output): Response begins at the scene

[1166] The user must take appropriate action to remove the cause of the anomaly and return the situation to safety. Specifically, the user must take appropriate action until the staff seizes the dangerous object and the police arrive. (Input): Start of on-site response (Output): Completion of anomaly response

[1167] Step 9:

[1168] Video archiving and post-processing

[1169] The server stores video of detected anomalies for later analysis and as evidence. Specific operations include storing the video data in a secure storage and making it accessible as needed. (Input): On-site response data (Output): Stored video of anomalies

[1170] The server uses the stored data to evaluate the system's performance and retrain the AI ​​model as necessary. Specifically, it analyzes past anomaly detection cases and uses them as a dataset to improve the model's accuracy. (Input): Stored anomaly video. (Output): Updated AI model.

[1171] (Application example 2)

[1172] 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."

[1173] Modern society requires rapid and appropriate responses to various abnormal situations that may occur in public places and educational institutions (for example, the appearance of a person armed with a knife or the abandonment of suspicious liquids). However, previous surveillance systems have focused on object recognition and behavioral recognition, making it difficult to combine emotion recognition for highly accurate anomaly detection. Furthermore, real-time notification capabilities have been limited, resulting in delayed responses when an abnormality occurs. This has sometimes led to insufficient safety in public places and educational institutions.

[1174] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1175] In this invention, the server includes a data acquisition means for collecting visual data for detecting anomalies, a data analysis means for detecting anomalies from the acquired visual data, a notification means for notifying anomalies detected by the data analysis means, an emotion recognition means for recognizing a user's emotion, and an integrated analysis means for determining a detection result by combining abnormal behavior and the user's emotion, thereby enabling highly accurate detection of anomalies and rapid response.

[1176] "Visual data" is digital data containing visual information that is collected by image capture devices such as cameras and sensors.

[1177] "Data acquisition means" is a general term for devices including cameras and various sensors installed to collect visual data.

[1178] "Data analysis means" includes software and hardware means for analyzing acquired visual data and detecting specific patterns or anomalies.

[1179] "Notification Methods" are the functions and devices used to notify personnel or law enforcement in real time of details of detected anomalies, including notifications in the form of email, SMS, in-app alerts, etc.

[1180] "Emotion recognition means" refers to algorithms and devices that analyze a user's facial expressions, voice, posture, etc., and identify their psychological state and emotions.

[1181] The "integrated analysis means" is a means for combining abnormal behavior and user emotion data to perform a comprehensive analysis and improve the accuracy of abnormality detection.

[1182] An "object recognition algorithm" is a computational procedure for detecting specific objects (e.g., knives, suspicious liquids) from video data.

[1183] A "behavior recognition algorithm" is a calculation procedure for analyzing the movements and actions of people in video data and identifying suspicious behavior.

[1184] The "storage means" is a device for storing video data of detected abnormalities for later analysis or use as evidence.

[1185] This invention is a system that quickly detects abnormalities and automates appropriate responses, improving safety in public places and educational institutions. This system is composed of a combination of data acquisition means, data analysis means, reporting means, emotion recognition means, and integrated analysis means.

[1186] Data Acquisition Method

[1187] The devices use cameras and sensors installed at each location to collect visual data, providing real-time footage of the entire public space or educational institution. For example, a camera installed on a train platform captures footage of the entire platform.

[1188] Data Analysis Methods

[1189] The server analyzes the acquired visual data. This analysis includes object recognition algorithms and behavior recognition algorithms. For example, it detects people holding knives or handling suspicious liquids. These object recognition and behavior recognition algorithms are implemented using frameworks such as OpenCV and Keras.

[1190] emotion recognition means

[1191] The server is equipped with emotion recognition means to analyze the user's facial expressions, voice, and posture, allowing it to identify a person's emotional state (e.g., anger, fear, sadness, etc.) The technologies used include deep learning models and supporting software such as TensorFlow and Keras models.

[1192] Integrated analysis method

[1193] The server combines abnormal behavior and user emotional data for comprehensive analysis, improving the accuracy of anomaly detection. This integrated analysis method further improves the accuracy of anomaly detection when, for example, the user is not only holding a knife but also has an angry expression.

[1194] Reporting means

[1195] The server provides a means of notification to notify personnel and police in real time of detected anomalies, including details of the anomalies, via email, SMS, in-app alerts, etc. Video data of the anomalies is stored in the cloud or local storage for later analysis or as evidence.

[1196] Specific examples

[1197] Cameras installed on station platforms capture passenger activity in real time. For example, the cameras capture a situation where there are many passengers on the platform during rush hour. The server receives the video data and first uses object recognition and behavior recognition algorithms to detect people holding knives or people behaving suspiciously. The server then analyzes the person's emotions from facial expressions captured by emotion recognition means. If a person holding a knife or with an angry expression is judged to be abnormal, an emergency notification is immediately sent to staff and the police. Staff receive the notification and rush to the scene.

[1198] Prompt Sentence Examples

[1199] It analyzes camera footage in real time to detect suspicious individuals and movements. It also uses facial recognition technology to analyze emotions and immediately notify staff and police if an abnormality is detected. Please list the hardware and software required for this system and explain how to implement it specifically.

[1200] This system is expected to enable rapid and highly accurate detection of abnormalities in public places and educational institutions, significantly improving safety.

[1201] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1202] Step 1:

[1203] The device uses cameras and sensors installed at each location to collect visual data. The cameras and sensors capture video data in real time and temporarily store the data in local storage. The input is the visual data from the camera, and the output is the temporarily stored video data.

[1204] Step 2:

[1205] The server preprocesses the video data received from the terminal. First, it filters the data to remove noise, and then converts it to the required resolution. This process prepares the data for the data analysis means to operate efficiently. The input is the temporarily stored video data, and the output is the filtered and resolution-converted video data.

[1206] Step 3:

[1207] The server analyzes the filtered video data. In this step, it applies object and behavior recognition algorithms to detect abnormal behaviors and objects. For example, it identifies people holding knives or suspicious liquids. The input is the filtered and resolution-converted video data, and the output is the detection result data indicating anomalies.

[1208] Step 4:

[1209] The server analyzes the user's emotions based on the detection result data. Using emotion recognition means, it analyzes the user's facial expressions, voice, posture, etc., and identifies emotions such as anger, sadness, and fear. The input is the detection result data and visual data regarding abnormal behavior and abnormal objects, and the output is the emotion recognition result data.

[1210] Step 5:

[1211] The server integrates the abnormal behavior and emotion recognition results to make a final anomaly determination. Specifically, if a person holding a knife also displays anger, it is judged to be highly abnormal. The input is abnormal behavior data and emotion recognition result data, and the output is the integrated anomaly detection result.

[1212] Step 6:

[1213] The server activates the notification mechanism based on the final anomaly detection results. Emergency notifications are sent to personnel and police in real time, and detailed reports are generated. Notification formats include email, SMS, and in-app alerts, and messages containing anomaly identification information are sent. The input is the consolidated anomaly detection results, and the output is emergency notifications and detailed reports.

[1214] Step 7:

[1215] The user receives notifications from the reporting means and responds promptly to the scene. Specifically, when a staff member receives a notification, he or she immediately heads to the scene to confirm and respond to the abnormality. The input is the emergency notification and detailed report, and the output is the actual abnormality response action.

[1216] Step 8:

[1217] The server stores video data when an anomaly is detected. The stored data is used for later analysis and as evidence. The input is the video data when an anomaly occurs, and the output is the data stored in the cloud or local storage.

[1218] This series of processing steps enables the system to detect anomalies with high accuracy and respond quickly, significantly improving safety in public places and educational institutions.

[1219] 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.

[1220] 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.

[1221] 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.

[1222] [Fourth embodiment]

[1223] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1224] 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.

[1225] 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).

[1226] 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.

[1227] 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.

[1228] 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).

[1229] 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.

[1230] 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.

[1231] 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.

[1232] 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.

[1233] 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.

[1234] 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.

[1235] 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."

[1236] This invention is a system that quickly detects abnormalities in public places and educational institutions and automates appropriate responses. This system combines video capture, analysis, and notification means to significantly improve safety at the site.

[1237] Program processing

[1238] 1. Data Collection

[1239] Terminal: Collects real-time video from surveillance cameras and sensors installed in various locations. For example, a camera installed on a station platform captures video of the entire platform.

[1240] Terminal: Temporarily stores collected video data and makes it available for immediate analysis. For example, video captured by a camera is temporarily stored in local storage.

[1241] 2. Data Preprocessing

[1242] Server: Filters the received video data to remove noise, converting it into a clear, easy-to-analyze image.

[1243] Server: Converts the filtered video data to the optimal resolution and prepares it for sending to the analysis means.

[1244] 3. Analysis for Anomaly Detection

[1245] Server: The filtered video data is fed into an AI model, where object and behavior recognition algorithms are applied, for example to detect people holding knives or handling suspicious liquids.

[1246] Server: Identifies specific abnormal behavior, such as detecting someone walking quickly or moving differently from the normal crowd.

[1247] 4. Identifying abnormal situations

[1248] Server: If an anomaly is detected based on the analysis results, a detailed report is generated, including the location, time, and type of anomaly detected.

[1249] Server: Transforms the report into a readable format and prepares it for notification.

[1250] 5. Notification Processing

[1251] Server: Based on the generated report, it immediately notifies officials and police via email, SMS, in-app alerts, etc.

[1252] Notification method: A notification is sent to the employee's device and an alert sound is generated to attract attention. For example, a notification is displayed on the computer of an employee working in the monitoring room.

[1253] 6. On-site response

[1254] User: Upon receiving the notification, staff or police rush to the scene and check the details of the abnormality. For example, staff rush to the scene while checking the surveillance camera footage.

[1255] User: Eliminate the cause of the anomaly and take appropriate action to return the situation to safety.

[1256] 7. Video storage and post-processing

[1257] Server: Stores video footage of detected anomalies for later analysis or as evidence. For example, it stores video footage of before and after the anomaly occurred in a specific format.

[1258] Server: Analyzes the stored video data and uses it to evaluate and improve system performance.

[1259] Specific examples

[1260] Example on a station platform:

[1261] Terminal: Cameras installed on station platforms capture passenger activity in real time. For example, cameras can capture the situation of large numbers of passengers on the platform during rush hour.

[1262] Server: Receives video data and analyzes abnormal behavior, such as detecting someone pulling out a knife in a crowd.

[1263] Notification method: If an abnormality is detected, an emergency notification will be sent to station staff and the police via email or SMS. For example, a staff member may receive a notification on their smartphone stating that "a person with a knife is in the center of the platform."

[1264] User: Staff rush to the scene and contact the police. For example, staff immediately head to the scene and handle the situation until the police arrive.

[1265] Thus, the present invention is a system that provides an embodiment for quickly detecting abnormalities in public places and educational institutions and taking appropriate measures.

[1266] The processing flow will be explained below.

[1267] Step 1:

[1268] Terminal: Collects real-time video from surveillance cameras and sensors installed in various locations. For example, a camera installed on a station platform captures video of the entire platform.

[1269] Step 2:

[1270] Terminal: Collected video data is temporarily stored in local storage, making it available for immediate analysis. For example, video captured by a camera is temporarily stored in a buffer.

[1271] Step 3:

[1272] Server: Performs filtering to remove noise from the received video data, for example, removing non-essential parts of the video data and converting it into a clear, easy-to-analyze format.

[1273] Step 4:

[1274] Server: Converts filtered video data to the optimal resolution and prepares it for sending to the analysis algorithm. Converts data into the format required by the analysis method.

[1275] Step 5:

[1276] Server: The filtered video data is fed into an AI model, where object and behavior recognition algorithms are applied, for example to detect people holding knives or handling suspicious liquids.

[1277] Step 6:

[1278] Server: Identifies specific abnormal behavior, such as people walking quickly or moving differently from the general crowd, thereby identifying suspicious individuals in real time.

[1279] Step 7:

[1280] Server: If an anomaly is detected based on the analysis results, a detailed report is generated, including the location, time, and type of anomaly detected.

[1281] Step 8:

[1282] Server: Transforms reports into a format that is easy to read and prepares them for notification, e.g., creating text and visual reports.

[1283] Step 9:

[1284] Server: Based on the generated report, it immediately notifies officials and police via email, SMS, in-app alerts, etc.

[1285] Step 10:

[1286] Terminal: A notification is sent to the employee's terminal and an alert sound is generated to attract attention. For example, a notification is displayed on the computer of an employee working in the monitoring room.

[1287] Step 11:

[1288] User: Upon receiving the notification, staff or police rush to the scene and check the details of the abnormality. For example, staff rush to the scene while checking the surveillance camera footage.

[1289] Step 12:

[1290] User: Eliminate the cause of the anomaly and take appropriate action to return the situation to safety. For example, staff will seize the dangerous material and coordinate with the police.

[1291] Step 13:

[1292] Server: Stores video footage of detected anomalies for later analysis or as evidence. For example, it stores video footage of before and after the anomaly occurred in a specific format.

[1293] Step 14:

[1294] Server: Analyzes the stored video data to evaluate and improve system performance, for example by retraining the AI ​​model based on past cases.

[1295] Example 1

[1296] 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."

[1297] Public places and educational institutions require the ability to quickly and accurately detect anomalies and automate appropriate responses. However, conventional systems lack the accuracy of anomaly detection and notification speed required, making manual monitoring and response necessary. Furthermore, the process of storing detected anomalies and post-analysis was often inefficient. A system that solves these problems and significantly improves safety is needed.

[1298] 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.

[1299] In this invention, the server includes a data acquisition means for collecting video data for detecting anomalies, a data preprocessing means for removing noise from the acquired video data and optimizing resolution, an analysis means for detecting anomalies from the video data processed by the data preprocessing means, a notification means for notifying anomalies detected by the analysis means, and a video storage means for storing video of the detected anomalies for use as later analysis or evidence. This enables rapid and accurate detection of anomalies, real-time notification and storage, and further subsequent analysis.

[1300] "Data acquisition means" refers to the devices and technologies that collect the video data necessary to detect abnormalities.

[1301] "Data pre-processing means" refers to technology or devices that perform processing to remove noise from acquired video data and optimize resolution.

[1302] "Analysis means" refers to a device that includes algorithms and techniques for detecting anomalies based on pre-processed video data.

[1303] "Notification means" refers to the technology or device used to notify relevant parties of abnormalities detected by the analysis means.

[1304] "Video storage means" refers to technology or equipment for storing video of detected abnormalities and using it for later analysis or as evidence.

[1305] "Object recognition algorithm" refers to the computational processes and techniques used to detect and identify objects based on video data.

[1306] "Behavior recognition algorithm" refers to the computational processes and technologies used to detect and identify behaviors based on video data and determine abnormal behavior.

[1307] "Emergency notification function" refers to technology and systems that notify relevant parties such as staff and police in real time of details of detected abnormalities.

[1308] The present invention relates to a system for quickly detecting abnormalities in public places and educational institutions and automating appropriate responses. The system includes a data acquisition unit, a data preprocessing unit, an analysis unit, a notification unit, and a video storage unit.

[1309] The system is programmed as follows:

[1310] First, surveillance cameras and other data acquisition methods installed at various locations collect video data in real time. For example, cameras installed on train platforms capture passenger movements.

[1311] The device then temporarily stores the collected video data and makes it available for immediate analysis using local storage.

[1312] The server then performs a filtering process to remove noise from the received video data, making the video clearer and easier to analyze. The filtered video data is then converted to an optimal resolution and prepared for transmission to the analysis means.

[1313] The analysis method inputs the filtered video data into an AI model and applies object and behavior recognition algorithms to detect, for example, people holding knives or handling suspicious liquids, as well as identify specific abnormal behaviors, such as people walking quickly or moving differently from the general crowd.

[1314] If the analysis results in an anomaly, the server generates a detailed report that includes the location, time, and type of anomaly detected, and converts the report into an easy-to-read format for notification.

[1315] The notification method will immediately notify employees and police based on the contents of the generated report. Notifications will be sent via email, SMS, in-app alerts, etc. For example, an emergency notification such as "There is a person with a knife on the station platform" will be sent to an employee's smartphone.

[1316] The users who receive the notification, i.e., officials or police, rush to the scene and check the details of the abnormality. Officials rush to the scene while checking the surveillance camera footage and take appropriate action to eliminate the cause of the abnormality.

[1317] Finally, the server stores the video of the detected anomaly for later analysis and as evidence. The stored video data can be analyzed to evaluate and improve the system's performance.

[1318] Prompt Sentence Examples

[1319] Below are some example prompts to input to the generative AI model:

[1320] "Please explain the most effective method for detecting abnormal behavior in public places."

[1321] "What are the key elements of a system that quickly detects anomalies and automatically notifies you?"

[1322] Thus, the present invention is a system that provides an embodiment for quickly detecting abnormalities in public places and educational institutions and taking appropriate measures.

[1323] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1324] Step 1: Data collection

[1325] Terminal: Surveillance cameras installed in various locations collect video data in real time. For example, cameras installed on train platforms capture passenger movements.

[1326] Input: Real-time video data from public places and educational institutions.

[1327] Output: Real-time video data.

[1328] Step 2: Temporarily save the video data

[1329] Terminal: Collected video data is temporarily stored in local storage. For example, video data captured by a camera is temporarily stored on a local drive so that it can be analyzed immediately.

[1330] Input: Real-time video data.

[1331] Output: Stored temporary video data.

[1332] Step 3: Filtering

[1333] Server: Filters the received video data to remove noise. By removing unnecessary noise and blur, the video becomes clear and easier to analyze.

[1334] Input: Stored temporary video data.

[1335] Output: Filtered video data.

[1336] Step 4: Optimize the resolution

[1337] Server: Converts the filtered video data to the appropriate resolution and prepares it for analysis. For example, compresses high-resolution video slightly to maintain analysis accuracy and speed.

[1338] Input: Filtered video data.

[1339] Output: Optimized resolution video data.

[1340] Step 5: Input to the AI ​​model

[1341] Server: The optimized resolution video data is input into the AI ​​model and an object recognition algorithm is applied to detect people and objects in the video data.

[1342] Input: Video data with optimized resolution.

[1343] Output: Object and person detection results.

[1344] Step 6: Applying the activity recognition algorithm

[1345] Server: Uses the results of the object recognition algorithm to apply the behavior recognition algorithm, for example, to detect people whose movements deviate from their normal behavioral patterns.

[1346] Input: Object and person detection results.

[1347] Output: Abnormal behavior detection results.

[1348] Step 7: Generate anomaly detection reports

[1349] Server: If the analysis results in an anomaly, a detailed report is generated, including the location, time, and specific details of the anomaly.

[1350] Input: Anomalous behavior detection results.

[1351] Output: Anomaly detection report.

[1352] Step 8: Convert the report to another format

[1353] Server: Converts the generated reports into a format that is easy to read, for example, creating reports in PDF or HTML format.

[1354] Input: Anomaly detection report.

[1355] Output: Formatted report.

[1356] Step 9: Call Emergency Services

[1357] Server: Based on the generated reports, it provides instant notifications to officials and police via email, SMS or in-app alerts.

[1358] Input: The formatted report.

[1359] Output: Emergency call notification.

[1360] Step 10: Displaying notifications

[1361] Notification method: An emergency notification will be displayed on the employee's terminal and an alert sound will be emitted to attract attention. For example, the notification will be displayed immediately on the computer of an employee on duty in the monitoring room.

[1362] Input: Emergency call notification.

[1363] Output: Notifications that appear on the device.

[1364] Step 11: Rush to the scene

[1365] User: Upon receiving the notification, staff and police rush to the scene to check the details of the abnormality. Staff head to the scene while checking the surveillance camera footage.

[1366] Input: Notifications that appear on your device.

[1367] Output: Check the details of the error.

[1368] Step 12: Dealing with abnormalities

[1369] User: Eliminate the cause of the anomaly and take appropriate action to return the situation to safety. For example, staff may apprehend the anomaly.

[1370] Input: Check the details of the abnormality.

[1371] Output: Safe state restored.

[1372] Step 13: Save the abnormal video

[1373] Server: Stores video footage of detected anomalies for later analysis and as evidence. Stores footage of several minutes before and after the anomaly occurs.

[1374] Input: Detailed confirmation of anomaly and video data.

[1375] Output: Archived footage of the anomaly.

[1376] Step 14: Performance evaluation and system improvement

[1377] Server: Analyzes the stored video data to evaluate and improve the system's performance. For example, it evaluates detection accuracy and notification speed, and improves the AI ​​model.

[1378] Input: Archived anomalous footage.

[1379] Output: System improvement measures and evaluation results.

[1380] (Application example 1)

[1381] 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."

[1382] In modern society, there is a demand for immediate detection and rapid response of abnormal behavior in public places and educational institutions. However, current systems can be slow to detect and notify abnormalities, or on-site staff may not be able to respond immediately. Furthermore, there is a lack of means to immediately communicate detailed information about abnormalities to staff visually and audibly, making it difficult to respond to abnormalities efficiently. The objective of the present invention is to solve these problems.

[1383] 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.

[1384] In this invention, the server includes a video acquisition means for collecting camera data for detecting abnormalities, an analysis means for detecting abnormalities from the acquired camera data, a notification means for notifying the user of the abnormality detected by the analysis means, and a wearable display means for immediately conveying the notification information from the notification means to the user's visual and auditory senses. This makes it possible to quickly detect abnormal behavior, notify staff in real time, and respond immediately on site.

[1385] "Video acquisition means" refers to equipment and software that collects camera data for detecting abnormalities.

[1386] "Analysis means" refers to algorithms and software for detecting anomalies in acquired camera data.

[1387] The "notification means" refers to a system and software for notifying staff and other relevant parties of abnormalities detected by the analysis means.

[1388] The "wearable display means" is a device for instantly conveying the notification information of the notification means to the user's visual and auditory senses.

[1389] An "object recognition algorithm" is a program for identifying specific objects (e.g., knives, suspicious liquids) from video data.

[1390] A "behavior recognition algorithm" is a program that analyzes people's movements from video data and detects abnormal behavior (e.g., suspicious movements).

[1391] The "emergency notification function" is a function that notifies staff and police in real time of detected abnormalities.

[1392] A "server" is a main computing device that performs a series of processes such as data collection, analysis, and notification.

[1393] "User" refers to the staff and other relevant parties who receive and respond to the notified abnormality information.

[1394] The system for realizing this invention is configured using the following hardware and software: The hardware used is a surveillance camera, a cloud server, and smart glasses, while the software used is an AI model for video analysis (e.g., YOLO, OpenPose), a library for video noise reduction and filtering (e.g., OpenCV), and a notification service API (e.g., Firebase Cloud Messaging, Twilio).

[1395] The server collects video data from surveillance cameras in real time, which is first pre-processed with a program to remove noise and perform filtering, improving the quality of the video data and making it easier to analyze.

[1396] The server then inputs the preprocessed video data into an AI model and applies object and behavior recognition algorithms, such as YOLO to detect knives and suspicious liquids and OpenPose to identify abnormal behaviors, enabling the detection of specific abnormal behaviors and objects in real time.

[1397] If the server detects an anomaly based on these analysis results, it generates a detailed report containing information such as the location, time, and type of anomaly. After the report is generated, an emergency notification is sent immediately to staff and relevant parties via the notification service API. Notifications are sent in multiple ways, including email, SMS, and in-app alerts.

[1398] The user, a staff member, receives notifications through the smart glasses. The wearable display immediately conveys the notification information visually and audibly, allowing the staff member to quickly grasp the situation on-site. The staff member can check the details of any abnormalities in real time and take prompt action.

[1399] Examples:

[1400] For example, if a surveillance camera installed on a station platform detects a person holding a knife in the crowd, the server will filter the footage and identify the person as holding a knife using the YOLO model. If an anomaly is detected, the server will generate a detailed report and send a notification to the employee's smart glasses via the notification service API, stating, "There is a person holding a knife in the center of the platform." The employee will receive the notification and confirm the visual and audio alert, allowing them to respond to the scene quickly.

[1401] Example prompt sentence:

[1402] "Analyze the footage from the center of the platform to detect if there is anyone holding a knife."

[1403] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1404] Step 1:

[1405] The device transmits real-time video data captured by the camera to the cloud server. The input is video data from the surveillance camera, and the output is temporarily stored video data. This process involves acquiring video data and immediately transferring it to the server.

[1406] Step 2:

[1407] The server performs filtering to remove noise from the received video data. The input is the temporarily stored video data, and the output is the noise-removed video data. This process uses the OpenCV library to process the data to improve the video quality.

[1408] Step 3:

[1409] The server converts the filtered video data to the optimal resolution and prepares it for transmission to the analysis means. The input is the noise-removed video data, and the output is video data converted to a resolution suitable for analysis. This process involves data calculations using high-performance image processing algorithms.

[1410] Step 4:

[1411] The server then inputs the converted video data into an AI model and applies object and action recognition algorithms. The input is video data suitable for analysis, and the output is detected anomaly information. This process uses YOLO and OpenPose for object and action recognition and data calculation.

[1412] Step 5:

[1413] The server determines anomalies based on the analysis results of the AI ​​model and generates a detailed report. The input is the detected anomaly information, and the output is a report summarizing the details of the anomaly. This process generates a document summarizing the type of anomaly, the location where it occurred, the time, etc.

[1414] Step 6:

[1415] Based on the generated report, the server sends emergency notifications to staff and related parties via the notification service API. The input is a detailed report, and the output is the sent notification. This process uses services such as Firebase Cloud Messaging and Twilio to send notifications immediately.

[1416] Step 7:

[1417] The user receives notifications through smart glasses and acknowledges visual and audible alerts. The input is the notification sent, and the output is the user's acknowledgement and immediate response. This process uses the wearable device to display information and provide alerts.

[1418] Step 8:

[1419] After checking the contents of the notification, the user quickly heads to the site, checks the details of the abnormality, and takes action. The input is the content of the notification and the situation at the site, and the output is the appropriate action to eliminate the cause of the abnormality. This process allows for a quick response and problem resolution at the site.

[1420] 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.

[1421] This invention is a system that quickly detects abnormalities in public places and educational institutions and automates appropriate responses. This system combines video acquisition means, analysis means, notification means, and an emotion engine that recognizes user emotions, significantly improving safety.

[1422] Program processing

[1423] 1. Data Collection

[1424] Terminal: Collects real-time video from surveillance cameras and sensors installed in various locations. For example, a camera installed on a station platform captures video of the entire platform.

[1425] Terminal: Collected video data is temporarily stored in local storage, making it available for immediate analysis. For example, video captured by a camera is temporarily stored in a buffer.

[1426] 2. Data Preprocessing

[1427] Server: Performs filtering to remove noise from the received video data, for example, removing non-essential parts of the video data and converting it into a clear, easy-to-analyze format.

[1428] Server: Converts filtered video data to the optimal resolution and prepares it for sending to the analysis algorithm. Converts data into the format required by the analysis method.

[1429] 3. Analysis for Anomaly Detection

[1430] Server: The filtered video data is fed into an AI model, where object and behavior recognition algorithms are applied, for example to detect people holding knives or handling suspicious liquids.

[1431] Server: Identifies specific abnormal behavior, such as people walking quickly or moving differently from the general crowd, thereby identifying suspicious individuals in real time.

[1432] 4. Analysis for Emotion Recognition

[1433] Server: Runs an emotion engine based on video data and analyzes the user's facial expressions, voice, posture, etc. For example, it analyzes emotions such as smile, anger, and sadness from camera footage.

[1434] Server: Emotion data determined by the emotion engine is supplied to the analysis means and used as data to improve the accuracy of anomaly detection. For example, if a person with an angry expression is holding a knife, the degree of anomaly is set high.

[1435] 5. Identifying abnormal situations

[1436] Server: If an anomaly is detected based on the analysis results, a detailed report is generated, including the location, time, and type of anomaly detected.

[1437] Server: Transforms reports into a format that is easy to read and prepares them for notification, e.g., creating text and visual reports.

[1438] 6. Notification Processing

[1439] Server: Based on the generated report, it immediately notifies officials and police via email, SMS, in-app alerts, etc.

[1440] Notification method: A notification is sent to the employee's device and an alert sound is generated to attract attention. For example, a notification is displayed on the computer of an employee working in the monitoring room.

[1441] 7. On-site response

[1442] User: Upon receiving the notification, staff or police rush to the scene and check the details of the abnormality. For example, staff rush to the scene while checking the surveillance camera footage.

[1443] User: Eliminate the cause of the anomaly and take appropriate action to return the situation to safety. For example, staff will seize the dangerous material and coordinate with the police.

[1444] 8. Video storage and post-processing

[1445] Server: Stores video footage of detected anomalies for later analysis or as evidence. For example, it stores video footage of before and after the anomaly occurred in a specific format.

[1446] Server: Analyzes the stored video data to evaluate and improve system performance, for example by retraining the AI ​​model based on past cases.

[1447] Specific examples

[1448] Example on a station platform:

[1449] Terminal: Cameras installed on station platforms capture passenger activity in real time. For example, cameras can capture the situation of large numbers of passengers on the platform during rush hour.

[1450] Server: After receiving the video data, the server first uses object recognition and behavior recognition algorithms to detect people holding a knife or making suspicious movements.Then, the server analyzes the person's emotions based on facial expressions captured by the emotion engine.

[1451] Server: A person holding a knife and with an angry expression is judged to be abnormal.

[1452] Notification method: If an abnormality is detected, an emergency notification will be sent to station staff and the police via email or SMS. For example, a staff member may receive a notification on their smartphone saying, "There is an angry person holding a knife in the middle of the platform."

[1453] User: Staff rush to the scene and contact the police. For example, staff immediately head to the scene and handle the situation until the police arrive.

[1454] In this way, by combining emotion engines, the present invention is a system that provides an embodiment for detecting anomalies with higher accuracy and taking prompt and appropriate action.

[1455] The processing flow will be explained below.

[1456] Step 1:

[1457] Terminal: Collects real-time video from surveillance cameras and sensors installed in various locations. For example, a camera installed on a station platform captures video of the entire platform.

[1458] Step 2:

[1459] Terminal: Collected video data is temporarily stored in local storage, making it available for immediate analysis. For example, video captured by a camera is temporarily stored in a buffer.

[1460] Step 3:

[1461] Server: Performs filtering to remove noise from the received video data, for example, removing non-essential parts of the video data and converting it into a clear, easy-to-analyze format.

[1462] Step 4:

[1463] Server: Converts filtered video data to the optimal resolution and prepares it for sending to the analysis algorithm. Converts data into the format required by the analysis method.

[1464] Step 5:

[1465] Server: The filtered video data is fed into an AI model, where object and behavior recognition algorithms are applied, for example to detect people holding knives or handling suspicious liquids.

[1466] Step 6:

[1467] Server: Identifies specific abnormal behavior, such as people walking quickly or moving differently from the general crowd, thereby identifying suspicious individuals in real time.

[1468] Step 7:

[1469] Server: Runs an emotion engine based on video data and analyzes the user's facial expressions, voice, posture, etc. For example, it analyzes emotions such as smile, anger, and sadness from camera footage.

[1470] Step 8:

[1471] Server: Emotion data determined by the emotion engine is supplied to the analysis means and used as data to improve the accuracy of anomaly detection. For example, if a person with an angry expression is holding a knife, the degree of anomaly is set high.

[1472] Step 9:

[1473] Server: If an anomaly is detected based on the analysis results, a detailed report is generated, including the location, time, and type of anomaly detected.

[1474] Step 10:

[1475] Server: Transforms reports into a format that is easy to read and prepares them for notification, e.g., creating text and visual reports.

[1476] Step 11:

[1477] Server: Based on the generated report, it immediately notifies officials and police via email, SMS, in-app alerts, etc.

[1478] Step 12:

[1479] Terminal: A notification is sent to the employee's terminal and an alert sound is generated to attract attention. For example, a notification is displayed on the computer of an employee working in the monitoring room.

[1480] Step 13:

[1481] User: Upon receiving the notification, staff or police rush to the scene and check the details of the abnormality. For example, staff rush to the scene while checking the surveillance camera footage.

[1482] Step 14:

[1483] User: Eliminate the cause of the anomaly and take appropriate action to return the situation to safety. For example, staff will seize the dangerous material and coordinate with the police.

[1484] Step 15:

[1485] Server: Stores video footage of detected anomalies for later analysis or as evidence. For example, it stores video footage of before and after the anomaly occurred in a specific format.

[1486] Step 16:

[1487] Server: Analyzes the stored video data to evaluate and improve system performance, for example by retraining the AI ​​model based on past cases.

[1488] Example 2

[1489] 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."

[1490] Public places and educational institutions are required to quickly detect suspicious individuals and abnormal behavior and respond appropriately. However, conventional systems have low anomaly detection accuracy and frequently generate false alarms. Furthermore, they are unable to respond quickly in emergencies, making it difficult to prevent serious damage. Therefore, there is a need to develop a new system that improves the accuracy of anomaly detection and enables quick and appropriate responses.

[1491] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes an acquisition means for collecting data, a preprocessing means for filtering the acquired data and converting its resolution, and an emotion recognition means for recognizing the user's emotion based on the preprocessed data. This makes it possible to analyze abnormal behavior and the user's emotion with high accuracy and to take prompt and appropriate action.

[1492] An "acquisition means" is a device or group of devices that collects data to detect abnormalities.

[1493] The "analysis means" refers to a processing device or algorithm for detecting anomalies based on the acquired data.

[1494] The "notification means" refers to a device or software that notifies relevant parties such as staff and police of abnormalities detected by the analysis means.

[1495] The "pre-processing means" refers to a processing device or algorithm that filters the acquired image data and converts it to an optimal resolution.

[1496] "Emotion recognition means" refers to an algorithm or device for analyzing the user's emotions based on acquired video data.

[1497] An "object recognition algorithm" is a program or processing method for analyzing video data and identifying specific objects (e.g., knives, suspicious liquids).

[1498] A "behavior recognition algorithm" is a program or processing method for analyzing video data and detecting specific behaviors (e.g., suspicious movements).

[1499] The "emergency notification function" is a function that sends real-time notifications to staff and police when an abnormality is detected.

[1500] This invention is a system that quickly detects abnormalities in public places and educational institutions and automates appropriate responses. This system combines video acquisition means, analysis means, notification means, and an emotion engine that recognizes user emotions, significantly improving safety.

[1501] The overall configuration of the system includes an "acquisition means" that collects data to detect abnormalities, a "preprocessing means" that filters and converts the resolution of the acquired data, an "analysis means" that analyzes the preprocessed data, a "notification means" that notifies the user of the analysis results, and an "emotion recognition means" that recognizes the user's emotions.

[1502] Explanation of program processing

[1503] Below is a detailed description of the process at each stage of the system.

[1504] Data collection

[1505] Terminal: Surveillance cameras and sensors installed in various locations collect video in real time. For example, a camera installed on a station platform captures video of the entire platform. These videos are temporarily stored in local storage.

[1506] Terminal: Collected video data is stored in buffer memory so that it can be analyzed immediately.

[1507] Data Preprocessing

[1508] Server: Performs filtering to remove noise from the received video data. Removes digital noise and adjusts the contrast and brightness of the video.

[1509] Server: The filtered video data is converted to the optimal resolution and prepared for transmission to the analysis algorithm. Data is compressed and format converted.

[1510] Analysis for anomaly detection

[1511] Server: The filtered video data is fed into the AI ​​model and an object recognition algorithm is applied. For example, the AI ​​model identifies knives, suspicious liquids, etc.

[1512] Server: Then, behavioral recognition algorithms are used to detect certain anomalous behaviors (e.g., rapid movements or suspicious movements), and behavioral patterns are analyzed to identify behaviors that differ from normal behavior.

[1513] Analysis for Emotion Recognition

[1514] Server: Runs an emotion engine based on video data, analyzing the user's facial expressions, voice, and posture. For example, it can read emotions such as smile, anger, and sadness from camera footage.

[1515] Server: Emotion data determined by the emotion engine is supplied to the analysis means to improve the accuracy of anomaly detection. The emotion data adjusts the anomaly parameter, increasing the accuracy of abnormal behavior.

[1516] Identifying abnormal situations

[1517] Server: If an anomaly is detected based on the analysis results, a detailed report is generated, including the location, time, and type of anomaly.

[1518] Server: Converts reports into a user-friendly format and prepares them for notification. Automatically generates reports in text and visual formats.

[1519] Notification Processing

[1520] Server: Based on the generated report, it immediately notifies officials and police, specifically by launching a system to send emails, SMS, and in-app alerts.

[1521] Notification method: A notification is sent to the employee's device, an alert sound is heard to attract attention, and a pop-up notification is displayed on the PC screen of the employee in the monitoring room.

[1522] On-site response

[1523] User: Upon receiving the notification, staff and police rush to the scene to check the details of the abnormality. They head to the scene while checking the surveillance camera footage.

[1524] User: Eliminate the cause of the anomaly and take appropriate action to return the situation to safety. Staff will seize the dangerous object and take appropriate measures until the police arrive.

[1525] Video archiving and post-processing

[1526] Server: Stores video of detected anomalies for later analysis and as evidence. Stores video data in a secure storage and makes it accessible when needed.

[1527] Server: Uses the stored data to evaluate system performance and retrain the AI ​​model if necessary. Analyzes past anomaly detection cases and uses them as a dataset to improve the model's accuracy.

[1528] Specific examples

[1529] Specific example on a station platform

[1530] Terminal: Cameras installed on station platforms capture passenger activity in real time. Images of the platform during rush hour are stored in buffer memory.

[1531] Server: Receives video data and uses object and behavior recognition algorithms to detect people holding knives or making suspicious movements. The emotion engine then analyzes the person's emotions based on their facial expressions.

[1532] Server: Identifies a person holding a knife and with an angry expression as abnormal and generates a detailed report.

[1533] Notification method: If an abnormality is detected, an emergency notification will be sent to station staff and police via email and SMS. Staff will also receive a notification on their smartphones stating that "an angry person holding a knife is in the middle of the platform."

[1534] User: Staff rush to the scene and contact the police. Staff immediately head to the scene and handle the situation until the police arrive.

[1535] Example prompts to input to the generative AI model

[1536] "Analyze abnormal behavior and emotions based on video data collected by cameras installed on station platforms. Specifically, you need to detect people holding knives or behaving abnormally, and analyze their emotions as well."

[1537] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1538] Step 1:

[1539] Data collection

[1540] The terminal collects video in real time from surveillance cameras and sensors installed at various locations. Specifically, cameras installed on the station platform continuously capture video of the entire platform. (Input): Surveillance camera video data (Output): Acquired video data (stored in buffer memory)

[1541] The terminal stores the collected video data in buffer memory, making it available for immediate analysis.

[1542] Step 2:

[1543] Data Preprocessing

[1544] The server performs filtering on the video data received from the buffer memory to remove noise. Specifically, it removes digital noise from the video and adjusts the contrast and brightness. (Input): Video data in the buffer memory (Output): Video data after filtering

[1545] The server converts the filtered video data to the optimal resolution and prepares it for transmission to the analysis algorithm. Specific operations include data compression and format conversion. (Input): Filtered video data (Output): Resolution-converted video data

[1546] Step 3:

[1547] object recognition

[1548] The server inputs the resolution-converted video data into the AI ​​model and applies an object recognition algorithm. Specifically, the AI ​​model identifies knives, suspicious liquids, etc. (Input): Resolution-converted video data (Output): Object recognition results (data identifying knives, suspicious objects, etc.)

[1549] Step 4:

[1550] behavioral recognition

[1551] The server uses a behavior recognition algorithm based on the results of object recognition to detect specific abnormal behavior. Specifically, the AI ​​model analyzes behavior patterns and identifies behaviors that differ from normal behavior. (Input): Object recognition results (Output): Behavior recognition results (data identifying abnormal behavior)

[1552] Step 5:

[1553] emotion recognition

[1554] The server runs an emotion engine based on the behavioral recognition data, analyzing the user's facial expressions, voice, and posture. Specific behaviors include smiles, anger, sadness, and other emotions read from the camera footage. (Input): Behavior recognition results (Output): Emotion recognition results (data with identified emotions)

[1555] The server supplies the emotion data determined by the emotion engine to the analysis means to improve the accuracy of anomaly detection. (Input): Emotion recognition results (Output): Data with anomaly level parameters adjusted

[1556] Step 6:

[1557] Anomaly detection and report generation

[1558] If the server determines that an anomaly has occurred based on the analysis results, it generates a detailed report. Specifically, it creates a report that clearly indicates the location, time, and type of anomaly. (Input): Data with anomaly parameter adjustments (Output): Anomaly detection report

[1559] The server converts the generated report into an easy-to-read format and prepares it for notification. Specific operations include automatically generating reports in text and visual formats. (Input): Anomaly detection report (Output): Report that can be notified

[1560] Step 7:

[1561] Notification Processing

[1562] The server immediately notifies officials and police based on the contents of the notifiable report. Specific actions include launching a system to send emails, SMS, and in-app alerts. (Input): Notifiable report (Output): Sent notification (email, SMS, etc.)

[1563] The notification method is that a notification is sent to the employee's terminal and an alert sound is emitted to attract their attention. Specific actions include a pop-up notification being displayed on the PC screen in the monitoring room. (Input): Sent notification (Output): Notification displayed on the employee's terminal

[1564] Step 8:

[1565] On-site response

[1566] Upon receiving the notification, the staff or police rush to the scene and check the details of the abnormality. Specific actions include heading to the scene while checking the video from the surveillance camera. (Input): Notification displayed on staff's device (Output): Response begins at the scene

[1567] The user must take appropriate action to remove the cause of the anomaly and return the situation to safety. Specifically, the user must take appropriate action until the staff seizes the dangerous object and the police arrive. (Input): Start of on-site response (Output): Completion of anomaly response

[1568] Step 9:

[1569] Video archiving and post-processing

[1570] The server stores video of detected anomalies for later analysis and as evidence. Specific operations include storing the video data in a secure storage and making it accessible as needed. (Input): On-site response data (Output): Stored video of anomalies

[1571] The server uses the stored data to evaluate the system's performance and retrain the AI ​​model as necessary. Specifically, it analyzes past anomaly detection cases and uses them as a dataset to improve the model's accuracy. (Input): Stored anomaly video. (Output): Updated AI model.

[1572] (Application example 2)

[1573] 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."

[1574] Modern society requires rapid and appropriate responses to various abnormal situations that may occur in public places and educational institutions (for example, the appearance of a person armed with a knife or the abandonment of suspicious liquids). However, previous surveillance systems have focused on object recognition and behavioral recognition, making it difficult to combine emotion recognition for highly accurate anomaly detection. Furthermore, real-time notification capabilities have been limited, resulting in delayed responses when an abnormality occurs. This has sometimes led to insufficient safety in public places and educational institutions.

[1575] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1576] In this invention, the server includes a data acquisition means for collecting visual data for detecting anomalies, a data analysis means for detecting anomalies from the acquired visual data, a notification means for notifying anomalies detected by the data analysis means, an emotion recognition means for recognizing a user's emotion, and an integrated analysis means for determining a detection result by combining abnormal behavior and the user's emotion, thereby enabling highly accurate detection of anomalies and rapid response.

[1577] "Visual data" is digital data containing visual information that is collected by image capture devices such as cameras and sensors.

[1578] "Data acquisition means" is a general term for devices including cameras and various sensors installed to collect visual data.

[1579] "Data analysis means" includes software and hardware means for analyzing acquired visual data and detecting specific patterns or anomalies.

[1580] "Notification Methods" are the functions and devices used to notify personnel or law enforcement in real time of details of detected anomalies, including notifications in the form of email, SMS, in-app alerts, etc.

[1581] "Emotion recognition means" refers to algorithms and devices that analyze a user's facial expressions, voice, posture, etc., and identify their psychological state and emotions.

[1582] The "integrated analysis means" is a means for combining abnormal behavior and user emotion data to perform a comprehensive analysis and improve the accuracy of abnormality detection.

[1583] An "object recognition algorithm" is a computational procedure for detecting specific objects (e.g., knives, suspicious liquids) from video data.

[1584] A "behavior recognition algorithm" is a calculation procedure for analyzing the movements and actions of people in video data and identifying suspicious behavior.

[1585] The "storage means" is a device for storing video data of detected abnormalities for later analysis or use as evidence.

[1586] This invention is a system that quickly detects abnormalities and automates appropriate responses, improving safety in public places and educational institutions. This system is composed of a combination of data acquisition means, data analysis means, reporting means, emotion recognition means, and integrated analysis means.

[1587] Data Acquisition Method

[1588] The devices use cameras and sensors installed at each location to collect visual data, providing real-time footage of the entire public space or educational institution. For example, a camera installed on a train platform captures footage of the entire platform.

[1589] Data Analysis Methods

[1590] The server analyzes the acquired visual data. This analysis includes object recognition algorithms and behavior recognition algorithms. For example, it detects people holding knives or handling suspicious liquids. These object recognition and behavior recognition algorithms are implemented using frameworks such as OpenCV and Keras.

[1591] emotion recognition means

[1592] The server is equipped with emotion recognition means to analyze the user's facial expressions, voice, and posture, allowing it to identify a person's emotional state (e.g., anger, fear, sadness, etc.) The technologies used include deep learning models and supporting software such as TensorFlow and Keras models.

[1593] Integrated analysis method

[1594] The server combines abnormal behavior and user emotional data for comprehensive analysis, improving the accuracy of anomaly detection. This integrated analysis method further improves the accuracy of anomaly detection when, for example, the user is not only holding a knife but also has an angry expression.

[1595] Reporting means

[1596] The server provides a means of notification to notify personnel and police in real time of detected anomalies, including details of the anomalies, via email, SMS, in-app alerts, etc. Video data of the anomalies is stored in the cloud or local storage for later analysis or as evidence.

[1597] Specific examples

[1598] Cameras installed on station platforms capture passenger activity in real time. For example, the cameras capture a situation where there are many passengers on the platform during rush hour. The server receives the video data and first uses object recognition and behavior recognition algorithms to detect people holding knives or people behaving suspiciously. The server then analyzes the person's emotions from facial expressions captured by emotion recognition means. If a person holding a knife or with an angry expression is judged to be abnormal, an emergency notification is immediately sent to staff and the police. Staff receive the notification and rush to the scene.

[1599] Prompt Sentence Examples

[1600] It analyzes camera footage in real time to detect suspicious individuals and movements. It also uses facial recognition technology to analyze emotions and immediately notify staff and police if an abnormality is detected. Please list the hardware and software required for this system and explain how to implement it specifically.

[1601] This system is expected to enable rapid and highly accurate detection of abnormalities in public places and educational institutions, significantly improving safety.

[1602] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1603] Step 1:

[1604] The device uses cameras and sensors installed at each location to collect visual data. The cameras and sensors capture video data in real time and temporarily store the data in local storage. The input is the visual data from the camera, and the output is the temporarily stored video data.

[1605] Step 2:

[1606] The server preprocesses the video data received from the terminal. First, it filters the data to remove noise, and then converts it to the required resolution. This process prepares the data for the data analysis means to operate efficiently. The input is the temporarily stored video data, and the output is the filtered and resolution-converted video data.

[1607] Step 3:

[1608] The server analyzes the filtered video data. In this step, it applies object and behavior recognition algorithms to detect abnormal behaviors and objects. For example, it identifies people holding knives or suspicious liquids. The input is the filtered and resolution-converted video data, and the output is the detection result data indicating anomalies.

[1609] Step 4:

[1610] The server analyzes the user's emotions based on the detection result data. Using emotion recognition means, it analyzes the user's facial expressions, voice, posture, etc., and identifies emotions such as anger, sadness, and fear. The input is the detection result data and visual data regarding abnormal behavior and abnormal objects, and the output is the emotion recognition result data.

[1611] Step 5:

[1612] The server integrates the abnormal behavior and emotion recognition results to make a final anomaly determination. Specifically, if a person holding a knife also displays anger, it is judged to be highly abnormal. The input is abnormal behavior data and emotion recognition result data, and the output is the integrated anomaly detection result.

[1613] Step 6:

[1614] The server activates the notification mechanism based on the final anomaly detection results. Emergency notifications are sent to personnel and police in real time, and detailed reports are generated. Notification formats include email, SMS, and in-app alerts, and messages containing anomaly identification information are sent. The input is the consolidated anomaly detection results, and the output is emergency notifications and detailed reports.

[1615] Step 7:

[1616] The user receives notifications from the reporting means and responds promptly to the scene. Specifically, when a staff member receives a notification, he or she immediately heads to the scene to confirm and respond to the abnormality. The input is the emergency notification and detailed report, and the output is the actual abnormality response action.

[1617] Step 8:

[1618] The server stores video data when an anomaly is detected. The stored data is used for later analysis and as evidence. The input is the video data when an anomaly occurs, and the output is the data stored in the cloud or local storage.

[1619] This series of processing steps enables the system to detect anomalies with high accuracy and respond quickly, significantly improving safety in public places and educational institutions.

[1620] 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.

[1621] 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.

[1622] 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.

[1623] 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.

[1624] 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.

[1625] 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.

[1626] 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).

[1627] 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, and motorcycles, 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.

[1628] 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."

[1629] 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.

[1630] 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).

[1631] 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.

[1632] 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 storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1633] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[1634] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1635] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1636] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1637] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1638] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1639] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1640] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1641] The following is further disclosed regarding the above embodiment.

[1642] (Claim 1)

[1643] an image acquisition means for collecting camera data for detecting anomalies;

[1644] an analysis means for detecting anomalies from the acquired camera data;

[1645] a notification means for notifying an abnormality detected by the analysis means;

[1646] A system including:

[1647] (Claim 2)

[1648] 2. The system of claim 1, wherein the analysis means includes object recognition algorithms and behavior recognition algorithms for detecting knives, suspicious liquids, and suspicious behavior.

[1649] (Claim 3)

[1650] 2. The system according to claim 1, wherein the notification means includes an emergency notification function for notifying officials and police in real time of details of the detected anomaly.

[1651] "Example 1"

[1652] (Claim 1)

[1653] a data acquisition means for collecting video data for detecting anomalies;

[1654] data pre-processing means for removing noise from the acquired image data and optimizing resolution;

[1655] analysis means for detecting anomalies from the video data processed by the data preprocessing means;

[1656] a notification means for notifying an abnormality detected by the analysis means;

[1657] A video storage means for storing video of detected abnormalities and using it for later analysis or as evidence;

[1658] A system including:

[1659] (Claim 2)

[1660] 2. The system of claim 1, wherein the analysis means includes an object recognition algorithm and an activity recognition algorithm.

[1661] (Claim 3)

[1662] The system of claim 1, wherein the notification means includes an emergency notification function for notifying officials and police in real time of details of anomaly detection based on the contents of the generated report.

[1663] "Application Example 1"

[1664] (Claim 1)

[1665] an image acquisition means for collecting camera data for detecting anomalies;

[1666] an analysis means for detecting anomalies from the acquired camera data;

[1667] a notification means for notifying an abnormality detected by the analysis means;

[1668] a wearable display means for instantly conveying the notification information of the notification means to the user's visual and auditory senses;

[1669] A system including:

[1670] (Claim 2)

[1671] 2. The system of claim 1, wherein the analysis means includes object recognition algorithms and behavior recognition algorithms for detecting knives, suspicious liquids, and suspicious behavior.

[1672] (Claim 3)

[1673] 2. The system according to claim 1, wherein the notification means includes an emergency notification function for notifying officials and police in real time of details of the detected anomaly.

[1674] "Example 2: Combining Emotion Engines"

[1675] (Claim 1)

[1676] an acquisition means for collecting data for detecting anomalies;

[1677] analysis means for detecting anomalies from the acquired data;

[1678] a notification means for notifying an abnormality detected by the analysis means;

[1679] a pre-processing means for filtering and converting the resolution of the image data obtained from the acquisition means;

[1680] emotion recognition means for recognizing the emotion of a user based on the data preprocessed by the analysis means;

[1681] The analysis means detects specific abnormal behavior using an object recognition algorithm and a behavior recognition algorithm;

[1682] means for the notification means to notify emergency personnel and police in real time;

[1683] A system including:

[1684] (Claim 2)

[1685] The system of claim 1, wherein the analysis means includes an object recognition algorithm and an action recognition algorithm for detecting knives, suspicious liquids, and suspicious behavior, and further includes emotion recognition means for analyzing the user's emotions.

[1686] (Claim 3)

[1687] 2. The system according to claim 1, wherein the notification means includes an emergency notification function for notifying officials and police in real time of details of the detected anomaly.

[1688] "Application example 2 when combining emotion engines"

[1689] (Claim 1)

[1690] data acquisition means for collecting visual data for detecting anomalies;

[1691] data analysis means for detecting anomalies from the acquired visual data;

[1692] a notification means for notifying an abnormality detected by the data analysis means;

[1693] emotion recognition means for recognizing an emotion of a user;

[1694] an integrated analysis means for combining the abnormal behavior and the user's emotion to determine the detection result;

[1695] ...

[1696] A system including:

[1697] (Claim 2)

[1698] 2. The system of claim 1, wherein the data analysis means includes object recognition algorithms and behavior recognition algorithms for detecting knives, suspicious liquids, and suspicious behavior, and the emotion recognition means further includes means for analyzing the user's facial expressions, voice, and posture.

[1699] (Claim 3)

[1700] The system of claim 1, wherein the reporting means further includes a reporting function and a reporting method for notifying staff and police in real time of details of the detected abnormality, and includes a storage means for storing video data of the abnormality for later analysis and use as evidence. [Explanation of symbols]

[1701] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. an image acquisition means for collecting camera data for detecting anomalies; an analysis means for detecting anomalies from the acquired camera data; a notification means for notifying an abnormality detected by the analysis means; A system including:

2. 2. The system of claim 1, wherein the analysis means includes object recognition algorithms and behavior recognition algorithms for detecting knives, suspicious liquids, and suspicious behavior.

3. 2. The system according to claim 1, wherein the notification means includes an emergency notification function for notifying officials and police in real time of details of the detected abnormality.

Images