System

Abstract

A system is provided.SOLUTION: A system comprising: image acquisition means for collecting image data of each vehicle; image analysis means for analyzing the image data to determine a congestion degree of each vehicle; information generation means for generating information indicating a congestion situation of each vehicle based on the congestion degree; and announcement means for announcing the congestion situation based on the information.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] Excessive crowding on commuter trains is not only a major source of stress for passengers, but also a safety issue. Recently, there have been a series of incidents of assaults on ordinary passengers by suspicious individuals armed with knives, and an increasing number of people are becoming anxious about using public transportation. Under these circumstances, there is a need to reduce stress for train users and ensure their safety. [Means for solving the problem]

[0005] The present invention is a system that includes an image acquisition means for collecting image data of each vehicle, an image analysis means for analyzing the image data to determine the congestion level of each vehicle, an information generation means for generating information indicating the congestion level of each vehicle based on the congestion level, and an announcement means for announcing the congestion level based on the information, thereby grasping the congestion level of commuter trains in real time and providing users with appropriate car selection information.The present invention also provides a system that further includes a feature detection means for detecting the features of a person passing through a ticket gate by referring to a criminal record database, a comparison means for comparing the features of the person detected by the feature detection means with data in the criminal record database, and a warning means for issuing a warning when a match is confirmed by the comparison means, thereby achieving early detection and response of suspicious persons and improving the safety of public transportation.

[0006] "Image acquisition means" refers to devices and their functions that are arranged to collect image data for each vehicle.

[0007] "Image analysis means" refers to technology or devices that analyze collected image data and determine the degree of congestion in each vehicle.

[0008] The "information generating means" is a function or device that generates information indicating the congestion status of each vehicle based on the analyzed congestion level.

[0009] "Announcement means" refers to a means for informing users of the congestion situation on trains and at stations based on the generated congestion information.

[0010] The "characteristic detection means" is a system or device for detecting the characteristics of a person passing through a ticket gate.

[0011] "Matching means" refers to technology or equipment that matches the characteristics of a detected person with data in a criminal record database to confirm a match.

[0012] The "warning means" is a system or device that issues a warning when a match is confirmed by the verification means.

[0013] "Crowding level" is an index that indicates the density of passengers and the degree of congestion in each vehicle.

[0014] "Announcement means" refers to a function or device that notifies users of the congestion situation based on the information. [Brief explanation of the drawings]

[0015] [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

[0016] 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.

[0017] First, the terms used in the following description will be explained.

[0018] 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).

[0019] 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.

[0020] 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.

[0021] 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.

[0022] 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."

[0023] [First embodiment]

[0024] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0025] 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.

[0026] 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).

[0027] 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.

[0028] 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.

[0029] 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.

[0030] 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.

[0031] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0032] 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.

[0033] 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.

[0034] 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.

[0035] 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."

[0036] The present invention provides a system for grasping the congestion status of commuter trains in real time and detecting suspicious individuals. The system includes an image acquisition unit, an image analysis unit, an information generation unit, an announcement unit, a feature detection unit, a comparison unit, and a warning unit.

[0037] Program processing and implementation form

[0038] 1. Image acquisition method

[0039] The terminal (in-train camera) periodically captures image data of each car. For example, a camera installed in each car captures an image of passengers every second and collects this image data.

[0040] 2. Image analysis methods

[0041] The server analyzes the image data sent from the device in real time using a trained machine learning model (e.g., convolutional neural network, CNN).The server uses the image data to determine the degree of congestion in each vehicle and quantifies the number and density of passengers in each vehicle.

[0042] 3. Information generation means

[0043] Based on the analyzed congestion level, the server generates information indicating the congestion status of each car (e.g., "Car 1 is 95% crowded," "Car 6 is 50% crowded," etc.). This information is stored in a database in real time.

[0044] 4. Announcement Methods

[0045] The server sends the generated congestion status information to the station announcement system. This allows station staff or automated announcements to provide users with information such as "Car 1 is very crowded. Car 6 is relatively empty." Users (commuters) can then select the car they want to board based on the announcements.

[0046] 5. Feature detection methods

[0047] The terminal (ticket gate system) captures a facial image of a person passing through the ticket gate, and the captured facial image data is sent to a server in real time.

[0048] 6. Verification Method

[0049] The server uses a facial recognition algorithm to extract features from the received facial image and compares it with a database of criminal records, which contains facial feature data of past criminals.

[0050] 7. Warning measures

[0051] If the server finds a match, it determines the person is suspicious and generates a warning. A warning message is displayed on the terminal (ticket gate system), and a warning notification is simultaneously sent to the station staff's terminal. This alert allows station staff to respond immediately and prevent crimes from occurring.

[0052] Example

[0053] For example, during rush hour on a certain weekday morning, car 1 is very crowded, but car 6 is relatively empty. In this case, image data taken by the device is sent to the server, which analyzes it and determines that "car 1 is 95% crowded" and "car 6 is 50% crowded." Based on this, an announcement is made that "car 6 is relatively empty." The user (commuter) can select car 6 based on this information to enjoy a comfortable commute.

[0054] Furthermore, if a suspicious person is detected while passing through a ticket gate, the server immediately sends an alert, which is displayed on the terminal (ticket gate system) and the station staff's terminal.Station staff can then respond quickly and prevent crimes from occurring.

[0055] Thus, the present invention is an effective means for alleviating congestion on commuter trains and improving safety.

[0056] The processing flow will be explained below.

[0057] Specific processing flow of the program

[0058] 1. Understanding congestion and making announcements

[0059] Step 1: Image acquisition

[0060] The terminal (in-train camera) captures real-time images from inside each car every second. This image data is compressed immediately after capture and sent to the server.

[0061] Step 2: Image reception and analysis

[0062] The server receives the image data sent from the terminal, immediately stores it in a database, and inputs it into the analysis module.

[0063] The server's image analysis module uses machine learning models (such as CNN) to analyze the congestion level of each vehicle, calculating the number of passengers, their density, the number of people standing, and other factors to quantify the level of congestion.

[0064] Step 3: Generate congestion information

[0065] Based on the analysis results, the server quantifies the degree of congestion in each vehicle and classifies it into categories such as "very crowded," "moderately crowded," or "empty" according to certain criteria.

[0066] Step 4: Generate announcement data

[0067] The server generates text data for announcements based on the generated congestion information. For example, "Car 1 is very crowded. Car 6 is relatively empty."

[0068] Step 5: Making the announcement

[0069] The server sends the generated text data to the station's announcement system, where it is processed by a speech synthesis module and an announcement is made.

[0070] Users (commuters) listen to the announcement and select the car they want to get in. For example, if they know that car 1 is crowded, they can choose car 6.

[0071] 2. Criminal Detection and Warning

[0072] Step 1: Image acquisition

[0073] The terminal (ticket gate system) takes facial images of all people passing through the ticket gate, and the captured image data is sent to the server in real time.

[0074] Step 2: Receiving and preprocessing face images

[0075] The server receives the facial image data sent from the ticket gate system, and immediately extracts facial feature points using a feature point extraction algorithm.

[0076] Step 3: Matching facial feature points

[0077] The server compares the minutiae data with a criminal record database. A matching algorithm calculates the degree of match between the minutiae and determines a match if the degree of match exceeds a certain threshold.

[0078] Step 4: Match detection and warning generation

[0079] If a match is found, the server generates alert data, which includes the suspect's characteristics and location.

[0080] Step 5: Sending an alert

[0081] The server sends the generated warning data to the terminal (ticket gate system) and the station staff's terminal. A warning message is displayed on the ticket gate system, and an alert is sent to the station staff's terminal.

[0082] Upon receiving the alert, station staff immediately take steps to deal with suspicious persons.

[0083] The above is the specific operation of each step. This system makes full use of image analysis and facial recognition technology to reduce congestion on commuter trains and improve safety.

[0084] Example 1

[0085] 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."

[0086] On current commuter trains, it is difficult to grasp the congestion status of each car in real time, which results in passengers often boarding cars that are more crowded. Also, if a suspicious person passes through the ticket gate, there are limited ways to quickly deal with the problem, which poses safety issues.

[0087] 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.

[0088] In this invention, the server includes image acquisition means for collecting image data of each vehicle, image analysis means for analyzing the image data to determine the degree of congestion of each vehicle, information generation means for generating information indicating the congestion status of each vehicle based on the congestion status, announcement means for announcing the congestion status based on the information, image acquisition means for taking facial images of people passing through the ticket gate, comparison means for analyzing the facial images to extract feature points and comparing them with a criminal record database, and warning means for generating and notifying warning data if a match is confirmed by the comparison means. This allows passengers to choose vehicles that are less crowded, improving safety.

[0089] The "image acquisition means" is a device for capturing and collecting image data of passengers and passersby inside trains and at ticket gates.

[0090] "Image analysis means" is a technology for analyzing acquired image data and extracting specific information (for example, congestion level or characteristic points).

[0091] An "information generation means" is a process or system for generating useful information (e.g., congestion status) from analyzed data.

[0092] "Announcement means" refers to audio and display devices used to inform passengers and other relevant parties of the situation based on the generated information.

[0093] The "matching means" is a system for comparing the feature points of the acquired facial image with an existing database to confirm a match.

[0094] The "warning means" is a mechanism for generating warning data and notifying relevant parties when a match is confirmed.

[0095] "Crowding level of each car" is an index that indicates the number and density of passengers in each car of a train.

[0096] A "criminal record database" is a database that registers facial images and feature points of people who have a criminal record.

[0097] "Feature detection means" is a technology for analyzing and extracting feature points from the facial image of a person passing through a ticket gate.

[0098] This invention is a system for understanding the congestion situation of commuter trains in real time and detecting suspicious individuals. This system operates through a series of processes: image data acquisition, analysis, and warning notification. Each function is explained in detail below.

[0099] Image Acquisition Method

[0100] The devices (in-train cameras and ticket gate system cameras) periodically capture image data of passengers and passersby and send it to a server. For example, high-resolution cameras (generic name) installed in each car and ticket gate system cameras (generic name) are used. These cameras automatically use a timer function to periodically capture images, store the data in temporary storage, and then send it to the server.

[0101] Image analysis methods

[0102] The server analyzes the received image data in real time. This analysis uses a pre-trained machine learning algorithm (e.g., convolutional neural network, generic name). Software used for the analysis includes TensorFlow (generic name). The received image data is processed by a server equipped with a high-performance GPU (generic name), which quantifies the number of passengers and their density to determine the level of congestion.

[0103] Information generation means

[0104] The server generates information showing the congestion status of each car based on the analyzed congestion level information. For example, information such as "Car 1 is 95% crowded" or "Car 6 is 50% crowded" is generated. This information is saved in a database in real time and kept up to date. MySQL (generic name) or similar is used as the database.

[0105] Announcement Method

[0106] The server sends the generated congestion status information to the announcement system within the station. This allows station staff or automated voice systems to provide users with information such as, "Car 1 is very crowded. Car 6 is relatively empty." Users (passengers) can choose their own car based on this information, allowing them to avoid crowds and enjoy a comfortable commute.

[0107] Feature detection and matching methods

[0108] The terminal (ticket gate system) takes a picture of the face of a person passing through the ticket gate and sends it to the server. The server analyzes the received face image using a facial recognition algorithm (e.g., OpenCV, generic name) and extracts feature points. The extracted feature points are then compared with a database of criminal records.

[0109] warning means

[0110] If the server finds a match, it determines the person is suspicious and generates a warning. The generated warning data is sent in real time to the ticket gate system and station staff terminals, where a warning message is displayed. This allows station staff to respond quickly and prevent crimes from occurring.

[0111] Specific examples

[0112] For example, during rush hour on a certain weekday morning, car 1 is very crowded, but car 6 is relatively empty. In this case, image data taken by the device is sent to the server, which analyzes it and determines that "car 1 is 95% crowded" and "car 6 is 50% crowded." Based on this, an announcement is made that "car 6 is relatively empty." The user (commuter) can select car 6 based on this information to enjoy a comfortable commute.

[0113] Furthermore, if a suspicious person is detected while passing through a ticket gate, the server immediately sends an alert, which is displayed on the ticket gate system and on the station staff's terminals, allowing station staff to respond quickly and prevent crimes from occurring.

[0114] Prompt Sentence Examples

[0115] "Please explain the system that monitors the congestion status of train cars in real time during the weekday morning rush hour and selects the appropriate car. Please also explain the process for detecting suspicious individuals at ticket gates and issuing warnings."

[0116] Thus, the present invention is a very effective means for alleviating congestion on commuter trains and improving safety.

[0117] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0118] Step 1: Image acquisition

[0119] The terminals (in-train cameras and ticket gate system cameras) capture image data of each car and people passing through the ticket gates. The input data are still images taken by the cameras, which are saved in temporary storage. Specifically, the cameras record images every second and temporarily save these images in memory. The saved image data is then sent to the server.

[0120] Step 2: Receiving image data

[0121] The server receives image data sent from the terminal. The input data is an image file taken by the terminal camera. The server receives this image data in real time and temporarily stores it in storage. This storage is used as the input data source for the next analysis process.

[0122] Step 3: Image analysis

[0123] The server analyzes the received image data. The input is the received image files, and the output is the congestion level of each car and facial feature data of people passing through the ticket gates. This analysis uses a trained machine learning algorithm (e.g., convolutional neural network, CNN). The server uses a GPU to process the images at high speed and extracts the number and density of passengers and facial feature points.

[0124] Step 4: Congestion assessment

[0125] The server quantifies the congestion level of each car based on the results of image analysis. The input is the number of passengers and density data obtained from image analysis, and the output is the congestion level of each car (e.g., "Car 1 is 95% crowded," "Car 6 is 50% crowded"). Based on this numerical data, the congestion level of each car is evaluated and information is generated.

[0126] Step 5: Information Generation

[0127] The server creates announcement data based on the generated congestion status information. The input data is the congestion assessment result, and the output is information in the form of a message to be sent to the announcement system. For example, information such as "Car 1 is very crowded. Car 6 is relatively empty" is generated.

[0128] Step 6: Announcement

[0129] The server sends the created announcement data to the announcement system in the station. The input data is the message information for the announcement, and the output is the information to be displayed on the station's speakers or displays. This allows station staff and automated voice systems to provide commuters with congestion information. Users (commuters) can select their train car based on this announcement information.

[0130] Step 7: Extract facial feature points

[0131] The server analyzes the facial image data of people passing through the ticket gate and extracts facial feature points. The input data is the facial image sent from the ticket gate system camera, and the output data is the extracted facial feature points. Here, the facial feature points are extracted using a machine learning algorithm (e.g., OpenCV).

[0132] Step 8: Face data matching

[0133] The server compares the extracted facial feature points with a database of criminal records. The input data is the facial feature points and known face data in the database, and the output data is the comparison result. If a match is confirmed as a result of the comparison, the person is judged to be suspicious.

[0134] Step 9: Generate warning data

[0135] The server generates warning data based on the matching results. The input data is the facial data matching result, and the output data is a warning message. If a suspicious person is detected, the server immediately generates warning data and sends the message to the terminal (ticket gate system or station staff terminal).

[0136] Step 10: Warning Notification

[0137] The terminals (ticket gate system and station staff terminal) receive the warning data sent from the server and display the warning message. The input data is the warning message from the server, and the output is the warning message displayed on the terminal display. Station staff can check this warning and take action immediately.

[0138] The above is a specific flow of each processing step in the system of the present invention, which makes it possible to grasp the congestion situation on commuter trains and detect suspicious individuals.

[0139] (Application example 1)

[0140] 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."

[0141] In modern manufacturing factories, robots are required to efficiently transport raw materials and products, but this is often hindered by congestion within the factory and the presence of obstacles. From a safety perspective, detecting suspicious objects and obstacles is also important, but existing systems are not able to adequately address this. There is a need for a new system that can solve these problems and achieve efficient and safe transportation.

[0142] 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.

[0143] In this invention, the server includes an image acquisition unit for collecting video data of each area, an image analysis unit for analyzing the video data to determine the congestion level of each area, an information generation unit for generating information indicating the congestion level of each area based on the congestion level, a feature detection unit for detecting suspicious objects and obstacles, and a warning unit for issuing a warning about suspicious objects and obstacles detected by the feature detection unit. This makes it possible to grasp the congestion level within the factory in real time and optimize the robot's transportation route. Furthermore, detecting suspicious objects and obstacles can improve safety.

[0144] "Image acquisition means" refers to a device such as a camera mounted on a robot or peripheral equipment, and is a means for collecting image data of an area.

[0145] The "image analysis means" is a means for analyzing collected image data to determine the degree of congestion in an area and the presence or absence of suspicious objects and obstacles. Specifically, it uses machine learning algorithms to analyze images.

[0146] The "information generating means" is a means for generating information indicating the situation of an area based on the degree of congestion obtained by the image analyzing means and the detected objects.

[0147] The "announcement means" is a means for communicating the generated congestion information to robots and operators in the factory by voice or display.

[0148] The "feature detection means" is a means for detecting suspicious objects and obstacles from the collected image data.

[0149] The "warning means" is a means for issuing a warning to notify the operator when a suspicious object or obstacle is detected.

[0150] A system for realizing the present invention is configured as follows.

[0151] First, cameras mounted on the robots and peripheral equipment collect video data of each area in real time. This is the image acquisition method. The collected video data is then sent to a server via a network.

[0152] The server receives the transmitted video data and analyzes it. For the analysis, it uses machine learning algorithms such as convolutional neural networks (CNN). This allows it to determine the degree of congestion in each area and the presence of obstacles and suspicious objects. This is the image analysis method.

[0153] Next, the server generates information indicating the congestion status and safety of each area based on the information obtained by the image analysis means. This is the information generation means.

[0154] The generated information is announced to the user (factory operator or robot). Instructions for the transport route are sent to the robot in real time via the announcement means. Specifically, instructions include "Please select a different route due to high congestion" or "Please proceed along this route."

[0155] Furthermore, the server uses a feature detection means to detect suspicious objects and obstacles. If any are detected, a warning means is activated to issue a warning to the operator. This allows for a quick response and improves factory safety.

[0156] Consider the following scenario as a specific example. For example, if a specific area in a factory is congested, images captured by a camera are sent to a server in real time. The server analyzes the images and determines that the congestion level in that area is 90% or higher, and sends an instruction to the robot to "select an alternative route." If a suspicious object is detected, an alert is immediately issued and a notification is sent to the operator.

[0157] In this way, this invention makes it possible to grasp the congestion situation in a factory in real time and optimize the robot's transportation route. It also improves safety by detecting suspicious objects and obstacles.

[0158] Examples of prompts include:

[0159] "Please tell me how to implement software that analyzes the congestion status and presence of obstacles in a factory in real time. The system uses cameras, robots, and servers, and uses Python, OpenCV, and Keras. Please explain with specific code examples."

[0160] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0161] Step 1:

[0162] The terminal (camera) collects video data from each area in the factory in real time. Specifically, the camera periodically captures images of the situation within the area and acquires the image data. The input is the video data captured by the camera, and the output is the data saved on the terminal.

[0163] Step 2:

[0164] The terminal (camera) transmits the collected video data to the server via the network. The server receives the data and prepares it for the next analysis step. The input is the video data, and the output is the data sent to the server.

[0165] Step 3:

[0166] The server preprocesses the received video data, specifically resizing and normalizing the image. The input is the received video data, and the output is the preprocessed image data.

[0167] Step 4:

[0168] The server inputs the preprocessed image data into a machine learning algorithm to analyze the level of congestion, obstacles, and the presence of suspicious objects. Here, a generative AI model such as a convolutional neural network (CNN) is used. The input is the preprocessed image data, and the output is the analysis results.

[0169] Step 5:

[0170] The server generates information based on the analysis results. Specifically, it generates information indicating the congestion status of each area and the presence of suspicious objects, and stores this information in a database. The input is the analysis results, and the output is the generated information data.

[0171] Step 6:

[0172] The server notifies the users (robots and operators) of the generated information through the announcement means. The robot is informed of the transport route and the operator of safety information. The input is the generated information data and the output is notification data.

[0173] Step 7:

[0174] The device (camera) uses the feature detection means to re-detect suspicious objects and obstacles, and if any abnormalities are found, it sends the data to the server again. The input is the newly captured video data, and the output is the re-analyzed data.

[0175] Step 8:

[0176] The server analyzes the received data again, and if an abnormality is found, it issues a warning to the operator using the warning means. The input is the reanalyzed data, and the output is a warning notification.

[0177] 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.

[0178] The present invention is a system that grasps the congestion status of commuter trains in real time, detects suspicious individuals, and recognizes user emotions to adjust announcements and displays. This system includes image acquisition means, image analysis means, information generation means, announcement means, feature detection means, comparison means, warning means, and an emotion engine.

[0179] Program processing and implementation form

[0180] 1. Understanding congestion and making announcements

[0181] The terminal (in-train camera) captures real-time images from inside each car every second. This image data is compressed immediately after capture and sent to the server.

[0182] The server receives the image data sent from the terminal and analyzes it using a trained machine learning model (such as CNN) to determine the degree of congestion in each vehicle. The analysis results are used to quantify the number and density of passengers in each vehicle.

[0183] The server generates information indicating the congestion status of each car based on the analyzed congestion level. For example, "Car 1 is very crowded" or "Car 6 is relatively empty." This information is saved in a database in real time.

[0184] The server generates text data for announcements based on the congestion information and sends it to the station's announcement system. This data is processed by a voice synthesis module, and the announcement is made. Users (commuters) can select the car they want to board based on this announcement.

[0185] 2. Criminal Detection and Warning

[0186] The terminal (ticket gate system) takes facial images of all people passing through the ticket gate and transmits them to the server in real time.

[0187] The server uses a facial recognition algorithm to compare the received facial image data with a database of criminal records. If a match is found, the person is deemed suspicious and a warning message is generated. The warning message is sent to the terminal (ticket gate system) and the station staff's terminal, where a warning message or alert is immediately displayed.

[0188] 3. User emotion recognition and announcement / display adjustment

[0189] The device (emotion engine) analyzes the commuter's facial expressions and recognizes the user's emotions. The emotion engine acquires facial expression data in real time and identifies emotions such as joy, surprise, sadness, and anger.

[0190] The server stores and analyzes the recognized emotion data. For example, if many users in a certain car show angry or dissatisfied expressions, the content of the announcement regarding the level of congestion can be adjusted as much as possible. An announcement that gives a sense of security, such as "Car 6 is relatively empty. Please use it," can be generated.

[0191] The server then sends the adjusted text data for the announcement to the station announcement system, which then makes the announcement. If necessary, the server can also provide the information to station staff terminals to prompt them to take appropriate action.

[0192] Example

[0193] For example, during the morning rush hour, car 1 may be very crowded, but car 6 may be relatively empty. In this case, image data from the device (in-train camera) is sent to the server, which determines that "car 1 is very crowded" and "car 6 is relatively empty."

[0194] Furthermore, if the device (emotion engine) analyzes the user's facial expressions and detects that many users are showing signs of dissatisfaction or stress, the server will adjust the announcement. An announcement will be made that takes into consideration emotions, such as "Car 6 is relatively empty. Please use it." Users (commuters) can enjoy a comfortable commute by hearing this announcement and heading to car 6.

[0195] Furthermore, if a suspicious person is detected when passing through a ticket gate, the server will immediately issue an alert and notify the terminal (ticket gate system) and the station staff's terminal, enabling a rapid response.

[0196] In this way, the present invention will alleviate congestion, adjust announcements based on emotion recognition, and detect suspicious individuals early, thereby significantly improving the safety and comfort of commuter trains.

[0197] The processing flow will be explained below.

[0198] Specific processing flow of the program

[0199] 1. Understanding congestion and making announcements

[0200] Step 1: Image acquisition

[0201] The terminal (in-train camera) captures real-time images from inside each car every second. The image data is compressed immediately after capture and sent to the server.

[0202] Step 2: Image reception and analysis

[0203] The server receives the image data sent from the terminal, stores it in a database, and inputs the data into the analysis module.

[0204] The server's image analysis module uses a trained machine learning model (such as CNN) to analyze image data and determine the level of congestion in each vehicle. The analysis calculates the number of passengers, their density, the number of people standing, and other factors to quantify the level of congestion.

[0205] Step 3: Generate congestion information

[0206] Based on the analysis results, the server quantifies the degree of congestion in each vehicle and classifies it into categories such as "very crowded," "moderately crowded," or "empty" according to certain criteria.

[0207] Step 4: Generate announcement data

[0208] The server generates text data for announcements based on the generated congestion information. For example, "Car 1 is very crowded. Car 6 is relatively empty."

[0209] Step 5: Making the announcement

[0210] The server sends the generated text data to the station's announcement system, where it is processed by a speech synthesis module and an announcement is made.

[0211] The user (commuter) listens to the announcement and selects the car they will board. They can choose car 6 to avoid the crowded car 1.

[0212] 2. Criminal Detection and Warning

[0213] Step 1: Image acquisition

[0214] The terminal (ticket gate system) takes facial images of all people passing through the ticket gate, and the captured image data is sent to the server in real time.

[0215] Step 2: Receiving and preprocessing face images

[0216] The server receives the facial image data sent from the ticket gate system and inputs it into a facial recognition algorithm that extracts feature points.

[0217] Step 3: Matching facial feature points

[0218] The server compares the extracted feature data with a database of criminal records. The matching algorithm calculates the degree of match between the feature points and determines a match if the degree of match exceeds a certain threshold.

[0219] Step 4: Match detection and warning generation

[0220] If a match is found, the server determines that the individual is suspicious and generates warning data, including the individual's characteristics and location.

[0221] Step 5: Sending an alert

[0222] The server sends the generated warning data to the terminal (ticket gate system) and the station staff's terminal. A warning message is displayed on the ticket gate system, and an alert is sent to the station staff's terminal.

[0223] Upon receiving the alert, station staff immediately take steps to deal with suspicious persons.

[0224] 3. User emotion recognition and announcement / display adjustment

[0225] Step 1: Acquire facial expression data

[0226] The device (emotion engine) captures the facial expressions of commuters in real time, and the captured data is sent to the emotion engine.

[0227] Step 2: Facial Expression Analysis

[0228] The server receives the facial expression data sent from the emotion engine and analyzes it using an emotion recognition algorithm, identifying emotions such as joy, surprise, sadness, and anger.

[0229] Step 3: Storing and analyzing sentiment data

[0230] The server stores the recognized emotion data and performs statistical analysis. If many users in a particular vehicle express anger or dissatisfaction, the emotion data is taken into account in addition to the congestion level.

[0231] Step 4: Adjust your announcement

[0232] The server adjusts the content of the announcement based on the emotional data, for example, "Car 6 is relatively empty. Please use it." to create a sense of security.

[0233] Step 5: Making the announcement

[0234] The server then sends the adjusted text data for the announcement to the station announcement system, which then makes the announcement. If necessary, the information is also provided to station staff's terminals, encouraging them to take appropriate action.

[0235] Users (commuters) can enjoy a more comfortable commute by listening to this announcement and following the emotionally sensitive instructions.

[0236] Example

[0237] For example, during the morning rush hour, car 1 may be very crowded, while car 6 may be relatively empty. In this case, image data from the device (in-train camera) is sent to the server, which then analyzes the data and concludes that "car 1 is very crowded" and "car 6 is relatively empty."

[0238] In addition, the device (emotion engine) analyzes the user's facial expressions, and if many users show expressions of dissatisfaction or stress, the server adjusts the announcement content and makes a reassuring announcement such as, "Car 6 is relatively empty. Please use it." Users (commuters) can hear this announcement and head to car 6 for a comfortable commute.

[0239] Furthermore, if a suspicious person is detected when passing through a ticket gate, the server will immediately issue an alert and notify the terminal (ticket gate system) and the station staff's terminal, allowing for a prompt response.

[0240] As such, the present invention is a system that can alleviate congestion, adjust announcements based on emotion recognition, and detect suspicious individuals early, thereby significantly improving the safety and comfort of commuter trains.

[0241] Example 2

[0242] 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."

[0243] There is a need for a system that can efficiently grasp the congestion situation on commuter trains, detect suspicious individuals early, recognize users' emotions, and provide a comfortable commuting environment. However, conventional technologies have difficulty satisfying all of these requirements simultaneously, and have not been able to sufficiently improve the safety and comfort of commuters.

[0244] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes: image acquisition means for collecting image data of each car; image analysis means for analyzing the image data to determine the congestion level of each car; information generation means for generating information indicating the congestion level of each car based on the congestion level; announcement means for announcing the congestion level based on the information; means for saving the congestion level information in a database in real time; feature detection means for detecting features of a person passing through a ticket gate by referring to a criminal record database; comparison means for comparing the features of the person detected by the feature detection means with data in the criminal record database; warning means for issuing a warning when a match is confirmed by the comparison means; means for transmitting the warning data to the ticket gate system and the station staff terminal; emotion analysis means for analyzing the user's facial expression and recognizing their emotion; and means for adjusting the content of the announcement based on the emotion data recognized by the emotion analysis means. This makes it possible to grasp the congestion level of commuter trains in real time and provide a comfortable commuting environment by detecting suspicious individuals and recognizing the user's emotion.

[0245] "Image acquisition means" refers to a device or function for collecting image data inside a vehicle.

[0246] The "image analysis means" refers to a device or function for analyzing collected image data and determining the degree of congestion inside the vehicle.

[0247] The "information generating means" refers to a device or function for generating information indicating the congestion status of each vehicle based on the analysis results.

[0248] The "announcement means" refers to a device or function for making announcements based on the generated congestion information.

[0249] The "means for storing in a database" refers to a device or function for storing the generated congestion status information in a database in real time.

[0250] "Characteristics detection means" refers to a device or function for detecting the characteristics of a person passing through a ticket gate.

[0251] The "matching means" is a device or function for matching the characteristics of a detected person with data in a criminal record database.

[0252] The "warning means" is a device or function for issuing a warning when a match is confirmed based on the comparison result.

[0253] The "means for transmitting to the ticket gate system and the station staff terminal" refers to a device or function for transmitting warning data to the ticket gate system and the station staff terminal.

[0254] The "emotion analysis means" is a device or function for analyzing the user's facial expression and recognizing the emotion.

[0255] The "means for adjusting the content of the announcement" refers to a device or function for adjusting the content of the announcement based on the recognized emotion data.

[0256] MODE FOR CARRYING OUT THE INVENTION

[0257] The present invention is a system that grasps the congestion status of commuter trains in real time, detects suspicious individuals, and recognizes user emotions to adjust announcements and displays. This system includes image acquisition means, image analysis means, information generation means, announcement means, feature detection means, comparison means, warning means, and an emotion engine.

[0258] 1. Understanding congestion and making announcements

[0259] The device (in-train camera) captures real-time video of each car every second. This image data is compressed and sent to the server. The server then analyzes this image data using a trained machine learning model (e.g., a convolutional neural network (CNN)) to determine the level of congestion in each car. The analysis results are then quantified as the number and density of passengers in each car. The server then uses this information to generate information indicating the congestion level of each car. For example, "Car 1 is very crowded" or "Car 6 is relatively empty." This information is then stored in a database in real time. The server then generates text data for announcements based on the congestion level information and sends it to the station's announcement system. This text data is processed by a speech synthesis module and announced in real time. Users (commuters) can use this information to choose a more comfortable car.

[0260] 2. Suspicious person detection and warning

[0261] The terminal (ticket gate system) takes a facial image of every person passing through the ticket gate and sends it to a server in real time. The server then uses a facial recognition algorithm (for example, dlib, an open-source facial recognition library) to compare the received facial image data with a database of criminal records. If the comparison results in a match with a suspicious individual, the server generates warning data. This warning data is sent to the terminal (ticket gate system) and the station staff's terminal, where a warning message or alert is immediately displayed.

[0262] 3. User emotion recognition and announcement / display adjustment

[0263] The terminal (emotion engine) analyzes commuters' facial expressions and recognizes their emotions. The emotion engine uses open-source facial expression recognition libraries (e.g., OpenCV and Dlib) to acquire facial expression data in real time and identify emotions such as joy, surprise, sadness, and anger. The server stores and analyzes the recognized emotion data. For example, if many users in a certain car show angry or dissatisfied expressions, the announcement content can be adjusted. Specifically, it can be changed to a more reassuring message such as, "Car 6 is relatively empty. Please use it." This adjusted text data for the announcement is also processed by the speech synthesis module and sent to the station announcement system. Furthermore, the server can provide information to station staff terminals as needed to prompt appropriate action.

[0264] Specific examples

[0265] For example, if car 1 is very crowded during the morning rush hour, image data captured by the terminal (in-train camera) is sent to the server. The server analyzes this data and determines that "car 1 is very crowded" or "car 6 is relatively empty." Additionally, if the terminal (emotion engine) analyzes the user's facial expressions and detects that many users are expressing dissatisfaction or stress, the server adjusts the content of the announcement. A reassuring announcement such as "car 6 is relatively empty. Please use it" is made. The user (commuter) can then refer to this announcement and head to car 6, enjoying a comfortable commute. Furthermore, if a suspicious person is detected at the ticket gate, the server immediately generates warning data and sends it to the terminal (ticket gate system) and the station staff's terminal. Thus, the present invention is a system that provides a safe and comfortable commuting environment through commuter train congestion status, emotion recognition, and suspicious person detection.

[0266] Prompt Sentence Examples

[0267] "Please explain the specific processing steps of a system that monitors the congestion status of each vehicle in real time during the morning rush hour, detects suspicious individuals, and reflects user emotions in announcements."

[0268] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0269] Step 1:

[0270] The terminal (train camera) captures real-time images from inside each car every second. The input is raw video data from inside each car, and the output is compressed image data. Specifically, the camera captures images from inside each car at high resolution and compresses the data in JPEG format or similar. The compressed image data is sent to the server.

[0271] Step 2:

[0272] The server receives image data sent from the device. The input is compressed image data, and the output is data ready for analysis. The server inputs the received image data into a trained machine learning model (e.g., a convolutional neural network (CNN)) to analyze the congestion level of each vehicle. Specifically, the server uses the CNN to detect people in the image and calculate the number of people.

[0273] Step 3:

[0274] The server quantifies the degree of congestion in each car based on the image analysis results. The input is the analyzed person detection data, and the output is numerical data indicating the degree of congestion in each car. The server generates specific congestion status information, such as "Car 1 is very crowded" or "Car 6 is relatively empty." The generated information is saved in a database in real time.

[0275] Step 4:

[0276] The server creates text data for announcements based on the generated congestion information. The input is congestion information, and the output is text data for announcements. This text data is then processed by a voice synthesis module and sent to the station announcement system. Specifically, announcements such as "Car 1 is very crowded" and "Car 6 is relatively empty" are made in real time.

[0277] Step 5:

[0278] The terminal (ticket gate system) takes facial images of all people passing through the ticket gate and sends them to the server in real time. The input is the facial image of the person passing through, and the output is the transmitted facial image data. Specifically, the ticket gate system uses a high-resolution camera to take facial images, compresses the data, and sends it to the server.

[0279] Step 6:

[0280] The server analyzes the received facial image data using a facial recognition algorithm and compares it with a database of criminal records. The input is the transmitted facial image data, and the output is the comparison result. If it matches a suspicious person, the server generates warning data and sends it to the ticket gate system and station staff terminals. Specifically, the server uses the dlib library to perform facial recognition and compares it with a database of criminal records.

[0281] Step 7:

[0282] The device (emotion engine) analyzes commuters' facial expressions and recognizes their emotions. The input is facial expression data from each vehicle, and the output is analyzed emotion data. The emotion engine uses open-source facial expression recognition libraries (e.g., OpenCV and Dlib). Specifically, the device acquires facial expression data in real time and identifies emotions such as joy, surprise, sadness, and anger.

[0283] Step 8:

[0284] The server stores and analyzes the recognized emotion data. The input is the analyzed emotion data, and the output is data for adjusting the content of announcements. For example, if many users in a certain car show expressions of anger or dissatisfaction, the server adjusts the announcement content to something more reassuring, such as "Car 6 is relatively empty. Please use it." The text data for the announcement is processed by a speech synthesis module and immediately sent to the station announcement system.

[0285] Specific examples

[0286] For example, if car 1 is extremely crowded during the morning rush hour, image data from the device (in-train camera) is sent to the server, which determines that "car 1 is extremely crowded" and "car 6 is relatively empty." Furthermore, the device (emotion engine) analyzes the user's facial expressions and detects that many users are showing signs of dissatisfaction or stress. The server then adjusts the announcement to "car 6 is relatively empty. Please use it." Based on this announcement, users can head to car 6 and enjoy a comfortable commute. Furthermore, if a suspicious person is detected while passing through the ticket gate, the server immediately issues an alert and notifies the terminal (ticket gate system) and the station staff's terminal. In this way, the system grasps the situation from multiple angles in real time and responds promptly.

[0287] (Application example 2)

[0288] 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."

[0289] Conventional commuter train systems lack the means to grasp congestion status in real time and provide appropriate announcements to passengers. Furthermore, there is a need for security measures such as adjusting announcements to reduce passenger stress and anxiety, and detecting suspicious individuals. This will improve passenger comfort and safety.

[0290] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes image acquisition means for collecting image data of each vehicle, image analysis means for analyzing the image data to determine the congestion level of each vehicle, information generation means for generating information indicating the congestion level of each vehicle based on the congestion level, announcement means for announcing the congestion level based on the information, emotion recognition means for recognizing emotions by analyzing facial expressions of passengers, and announcement adjustment means for adjusting the content of announcements based on the emotions recognized by the emotion recognition means. This makes it possible to grasp the congestion level in real time, detect suspicious individuals early, and adjust announcements taking passenger emotions into consideration.

[0291] The "image acquisition means" is a mechanism that collects image data inside the vehicle using sensors such as cameras installed in each vehicle.

[0292] The "image analysis means" is a device or software for analyzing collected image data and determining the degree of congestion in each vehicle.

[0293] The "information generating means" is a mechanism that generates information indicating the congestion state of each vehicle based on the analyzed congestion degree.

[0294] The "announcement means" is a system that announces the congestion situation to passengers based on the generated congestion situation information.

[0295] The "emotion recognition means" is a device or software that analyzes the facial expressions of passengers captured by cameras inside the vehicle and recognizes their emotions.

[0296] The "announcement adjustment means" is a system that adjusts the content and manner of announcements based on the recognized emotions.

[0297] In this invention, in order to improve the comfort and safety of self-driving vehicles, a system is constructed that grasps the congestion situation in each vehicle and the emotions of passengers in real time, and makes appropriate announcements and displays based on that information.

[0298] The server includes the following means:

[0299] 1. Image acquisition method:

[0300] Real-time images are acquired using cameras installed inside the vehicle.

[0301] Cameras are installed at various locations within the vehicle to provide comprehensive coverage of the situation within each vehicle.

[0302] 2. Image analysis methods:

[0303] The acquired image data is compressed and transmitted to the server.

[0304] The server analyzes the image data using machine learning algorithms, such as convolutional neural networks (CNNs), to determine the level of congestion in each vehicle.

[0305] 3. Information generation means:

[0306] Based on the analysis results, information indicating the congestion status of each car is generated, such as "Car 1 is very crowded" or "Car 6 is relatively empty."

[0307] The generated information is stored in a database on the server in real time.

[0308] 4. Announcement Method:

[0309] Text data for announcements is generated based on the information indicating the congestion state.

[0310] The generated text data is converted into speech via a speech synthesis module and transmitted to the announcement system inside the train or station.

[0311] 5. Emotion recognition means:

[0312] The server uses footage from cameras inside the vehicle to analyze passengers' facial expressions in real time and recognize emotions such as joy, anger, sadness, and surprise.

[0313] Emotion recognition uses techniques such as convolutional neural networks (CNNs) for emotion classification.

[0314] 6. Announcement Coordination Methods:

[0315] Adjust the content and manner of announcements based on the recognized emotional data.

[0316] For example, if many passengers look dissatisfied, an announcement will be made saying, "The next car is available, so please use it."

[0317] As a concrete example, imagine a situation where car 1 is very crowded during the morning rush hour, but car 6 is relatively empty. Image data captured by the camera is analyzed by the server, and information such as "car 1 is very crowded" and "car 6 is relatively empty" is generated. This information is provided to passengers as a voice announcement. Furthermore, if the passenger's facial expression indicates dissatisfaction, the server will make a more considerate announcement.

[0318] An example of a prompt is shown below.

[0319] "We will develop a system that analyzes the congestion situation and passenger emotions in real time from camera footage of autonomous vehicles, and adjusts the content of announcements and displays based on that analysis. For example, if the congestion level is high and passengers are expressing dissatisfaction, the system will announce that the next car is available, improving passenger comfort."

[0320] In this way, it is possible to realize a system that enhances the comfort and safety of autonomous vehicles and improves passenger satisfaction.

[0321] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0322] Step 1:

[0323] The terminal (in-car camera) captures real-time images from inside the car. The camera is installed in a specific position inside each car and captures the overall congestion situation and the facial expressions of passengers. The camera captures image data every second, compresses this data, and sends it to the server.

[0324] Input: Real-time video from inside the vehicle

[0325] Output: Compressed image data

[0326] Step 2:

[0327] The server receives the image data sent from the device. Each received image is analyzed using a machine learning algorithm. Specifically, a convolutional neural network (CNN) model is used to determine the degree of vehicle congestion.

[0328] Input: Compressed image data

[0329] Action: Data extraction and image analysis

[0330] Output: Quantified congestion data

[0331] Step 3:

[0332] The server generates information indicating the congestion status of each car based on the analysis results. For example, if the analysis shows that car 1 is highly crowded, it generates information such as "car 1 is very crowded."

[0333] Input: Data quantifying congestion

[0334] Action: Generate congestion information

[0335] Output: Information showing the occupancy status of each vehicle

[0336] Step 4:

[0337] The server then generates text data for announcements based on the congestion status, which is then converted into speech by a speech synthesis module and sent to the announcement system.

[0338] Input: Information indicating the occupancy status of each vehicle

[0339] Operation: Generation of text data for announcements and speech synthesis

[0340] Output: Announcement audio data

[0341] Step 5:

[0342] The terminals (speakers inside the train and in stations) receive the announcement audio data sent from the server and make the actual announcements, allowing passengers to understand the congestion situation in real time.

[0343] Input: Announcement audio data

[0344] Action: Make an announcement

[0345] Output: Announcement voice

[0346] Step 6:

[0347] The terminal (in-vehicle camera) captures facial expression data of passengers. Similarly, image data is captured every second, compressed, and sent to the server.

[0348] Input: Real-time video of passenger facial expressions

[0349] Output: Compressed facial expression data

[0350] Step 7:

[0351] The server receives the facial expression data and uses a machine learning model for emotion recognition (e.g., emotion classification model) to recognize emotions such as joy, anger, sadness, and surprise.

[0352] Input: Compressed facial expression data

[0353] Action: Facial expression analysis and emotion recognition

[0354] Output: Emotion data for each passenger

[0355] Step 8:

[0356] The server then adjusts the announcement content based on the recognized emotion data. For example, if many passengers show expressions of dissatisfaction or stress, the announcement will be changed to one that promotes a sense of security, such as "The next car is available, so please use it."

[0357] Input: Emotion data for each passenger

[0358] Action: Adjust announcement content

[0359] Output: Adjusted text data for announcement

[0360] Step 9:

[0361] The server converts the adjusted text data for announcement into voice again using the voice synthesis module and transmits it to the announcement system.

[0362] Input: Adjusted text data for announcement

[0363] Operation: Generation of announcement voice data and speech synthesis

[0364] Output: Adjusted announcement audio data

[0365] Step 10:

[0366] The terminals (speakers inside the train and at stations) receive the adjusted announcement audio data sent from the server and make announcements based on it.

[0367] Input: Adjusted announcement audio data

[0368] Action: Announcement is performed after adjustment

[0369] Output: Adjusted announcement audio

[0370] In this way, a system is built that can grasp congestion status in real time and make appropriate announcements based on that information. In addition, by adjusting the content of announcements to take passengers' emotions into consideration, a comfortable riding environment is provided.

[0371] 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.

[0372] 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.

[0373] 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.

[0374] [Second embodiment]

[0375] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0376] 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.

[0377] 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).

[0378] 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.

[0379] 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.

[0380] 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).

[0381] 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.

[0382] 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.

[0383] 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.

[0384] 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.

[0385] 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.

[0386] 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."

[0387] The present invention provides a system for grasping the congestion status of commuter trains in real time and detecting suspicious individuals. The system includes an image acquisition unit, an image analysis unit, an information generation unit, an announcement unit, a feature detection unit, a comparison unit, and a warning unit.

[0388] Program processing and implementation form

[0389] 1. Image acquisition method

[0390] The terminal (in-train camera) periodically captures image data of each car. For example, a camera installed in each car captures an image of passengers every second and collects this image data.

[0391] 2. Image analysis methods

[0392] The server analyzes the image data sent from the device in real time using a trained machine learning model (e.g., convolutional neural network, CNN).The server uses the image data to determine the degree of congestion in each vehicle and quantifies the number and density of passengers in each vehicle.

[0393] 3. Information generation means

[0394] Based on the analyzed congestion level, the server generates information indicating the congestion status of each car (e.g., "Car 1 is 95% crowded," "Car 6 is 50% crowded," etc.). This information is stored in a database in real time.

[0395] 4. Announcement Methods

[0396] The server sends the generated congestion status information to the station announcement system. This allows station staff or automated announcements to provide users with information such as "Car 1 is very crowded. Car 6 is relatively empty." Users (commuters) can then select the car they want to board based on the announcements.

[0397] 5. Feature detection methods

[0398] The terminal (ticket gate system) captures a facial image of a person passing through the ticket gate, and the captured facial image data is sent to a server in real time.

[0399] 6. Verification Method

[0400] The server uses a facial recognition algorithm to extract features from the received facial image and compares it with a database of criminal records, which contains facial feature data of past criminals.

[0401] 7. Warning measures

[0402] If the server finds a match, it determines the person is suspicious and generates a warning. A warning message is displayed on the terminal (ticket gate system), and a warning notification is simultaneously sent to the station staff's terminal. This alert allows station staff to respond immediately and prevent crimes from occurring.

[0403] Example

[0404] For example, during rush hour on a certain weekday morning, car 1 is very crowded, but car 6 is relatively empty. In this case, image data taken by the device is sent to the server, which analyzes it and determines that "car 1 is 95% crowded" and "car 6 is 50% crowded." Based on this, an announcement is made that "car 6 is relatively empty." The user (commuter) can select car 6 based on this information to enjoy a comfortable commute.

[0405] Furthermore, if a suspicious person is detected while passing through a ticket gate, the server immediately sends an alert, which is displayed on the terminal (ticket gate system) and the station staff's terminal.Station staff can then respond quickly and prevent crimes from occurring.

[0406] Thus, the present invention is an effective means for alleviating congestion on commuter trains and improving safety.

[0407] The processing flow will be explained below.

[0408] Specific processing flow of the program

[0409] 1. Understanding congestion and making announcements

[0410] Step 1: Image acquisition

[0411] The terminal (in-train camera) captures real-time images from inside each car every second. This image data is compressed immediately after capture and sent to the server.

[0412] Step 2: Image reception and analysis

[0413] The server receives the image data sent from the terminal, immediately stores it in a database, and inputs it into the analysis module.

[0414] The server's image analysis module uses machine learning models (such as CNN) to analyze the congestion level of each vehicle, calculating the number of passengers, their density, the number of people standing, and other factors to quantify the level of congestion.

[0415] Step 3: Generate congestion information

[0416] Based on the analysis results, the server quantifies the degree of congestion in each vehicle and classifies it into categories such as "very crowded," "moderately crowded," or "empty" according to certain criteria.

[0417] Step 4: Generate announcement data

[0418] The server generates text data for announcements based on the generated congestion information. For example, "Car 1 is very crowded. Car 6 is relatively empty."

[0419] Step 5: Making the announcement

[0420] The server sends the generated text data to the station's announcement system, where it is processed by a speech synthesis module and an announcement is made.

[0421] Users (commuters) listen to the announcement and select the car they want to get in. For example, if they know that car 1 is crowded, they can choose car 6.

[0422] 2. Criminal Detection and Warning

[0423] Step 1: Image acquisition

[0424] The terminal (ticket gate system) takes facial images of all people passing through the ticket gate, and the captured image data is sent to the server in real time.

[0425] Step 2: Receiving and preprocessing face images

[0426] The server receives the facial image data sent from the ticket gate system, and immediately extracts facial feature points using a feature point extraction algorithm.

[0427] Step 3: Matching facial feature points

[0428] The server compares the minutiae data with a criminal record database. A matching algorithm calculates the degree of match between the minutiae and determines a match if the degree of match exceeds a certain threshold.

[0429] Step 4: Match detection and warning generation

[0430] If a match is found, the server generates alert data, which includes the suspect's characteristics and location.

[0431] Step 5: Sending an alert

[0432] The server sends the generated warning data to the terminal (ticket gate system) and the station staff's terminal. A warning message is displayed on the ticket gate system, and an alert is sent to the station staff's terminal.

[0433] Upon receiving the alert, station staff immediately take steps to deal with suspicious persons.

[0434] The above is the specific operation of each step. This system makes full use of image analysis and facial recognition technology to reduce congestion on commuter trains and improve safety.

[0435] Example 1

[0436] 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."

[0437] On current commuter trains, it is difficult to grasp the congestion status of each car in real time, which results in passengers often boarding cars that are more crowded. Also, if a suspicious person passes through the ticket gate, there are limited ways to quickly deal with the problem, which poses safety issues.

[0438] 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.

[0439] In this invention, the server includes image acquisition means for collecting image data of each vehicle, image analysis means for analyzing the image data to determine the degree of congestion of each vehicle, information generation means for generating information indicating the congestion status of each vehicle based on the congestion status, announcement means for announcing the congestion status based on the information, image acquisition means for taking facial images of people passing through the ticket gate, comparison means for analyzing the facial images to extract feature points and comparing them with a criminal record database, and warning means for generating and notifying warning data if a match is confirmed by the comparison means. This allows passengers to choose vehicles that are less crowded, improving safety.

[0440] The "image acquisition means" is a device for capturing and collecting image data of passengers and passersby inside trains and at ticket gates.

[0441] "Image analysis means" is a technology for analyzing acquired image data and extracting specific information (for example, congestion level or characteristic points).

[0442] An "information generation means" is a process or system for generating useful information (e.g., congestion status) from analyzed data.

[0443] "Announcement means" refers to audio and display devices used to inform passengers and other relevant parties of the situation based on the generated information.

[0444] The "matching means" is a system for comparing the feature points of the acquired facial image with an existing database to confirm a match.

[0445] The "warning means" is a mechanism for generating warning data and notifying relevant parties when a match is confirmed.

[0446] "Crowding level of each car" is an index that indicates the number and density of passengers in each car of a train.

[0447] A "criminal record database" is a database that registers facial images and feature points of people who have a criminal record.

[0448] "Feature detection means" is a technology for analyzing and extracting feature points from the facial image of a person passing through a ticket gate.

[0449] This invention is a system for understanding the congestion situation of commuter trains in real time and detecting suspicious individuals. This system operates through a series of processes: image data acquisition, analysis, and warning notification. Each function is explained in detail below.

[0450] Image Acquisition Method

[0451] The devices (in-train cameras and ticket gate system cameras) periodically capture image data of passengers and passersby and send it to a server. For example, high-resolution cameras (generic name) installed in each car and ticket gate system cameras (generic name) are used. These cameras automatically use a timer function to periodically capture images, store the data in temporary storage, and then send it to the server.

[0452] Image analysis methods

[0453] The server analyzes the received image data in real time. This analysis uses a pre-trained machine learning algorithm (e.g., convolutional neural network, generic name). Software used for the analysis includes TensorFlow (generic name). The received image data is processed by a server equipped with a high-performance GPU (generic name), which quantifies the number of passengers and their density to determine the level of congestion.

[0454] Information generation means

[0455] The server generates information showing the congestion status of each car based on the analyzed congestion level information. For example, information such as "Car 1 is 95% crowded" or "Car 6 is 50% crowded" is generated. This information is saved in a database in real time and kept up to date. MySQL (generic name) or similar is used as the database.

[0456] Announcement Method

[0457] The server sends the generated congestion status information to the announcement system within the station. This allows station staff or automated voice systems to provide users with information such as, "Car 1 is very crowded. Car 6 is relatively empty." Users (passengers) can choose their own car based on this information, allowing them to avoid crowds and enjoy a comfortable commute.

[0458] Feature detection and matching methods

[0459] The terminal (ticket gate system) takes a picture of the face of a person passing through the ticket gate and sends it to the server. The server analyzes the received face image using a facial recognition algorithm (e.g., OpenCV, generic name) and extracts feature points. The extracted feature points are then compared with a database of criminal records.

[0460] warning means

[0461] If the server finds a match, it determines the person is suspicious and generates a warning. The generated warning data is sent in real time to the ticket gate system and station staff terminals, where a warning message is displayed. This allows station staff to respond quickly and prevent crimes from occurring.

[0462] Specific examples

[0463] For example, during rush hour on a certain weekday morning, car 1 is very crowded, but car 6 is relatively empty. In this case, image data taken by the device is sent to the server, which analyzes it and determines that "car 1 is 95% crowded" and "car 6 is 50% crowded." Based on this, an announcement is made that "car 6 is relatively empty." The user (commuter) can select car 6 based on this information to enjoy a comfortable commute.

[0464] Furthermore, if a suspicious person is detected while passing through a ticket gate, the server immediately sends an alert, which is displayed on the ticket gate system and on the station staff's terminals, allowing station staff to respond quickly and prevent crimes from occurring.

[0465] Prompt Sentence Examples

[0466] "Please explain the system that monitors the congestion status of train cars in real time during the weekday morning rush hour and selects the appropriate car. Please also explain the process for detecting suspicious individuals at ticket gates and issuing warnings."

[0467] Thus, the present invention is a very effective means for alleviating congestion on commuter trains and improving safety.

[0468] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0469] Step 1: Image acquisition

[0470] The terminals (in-train cameras and ticket gate system cameras) capture image data of each car and people passing through the ticket gates. The input data are still images taken by the cameras, which are saved in temporary storage. Specifically, the cameras record images every second and temporarily save these images in memory. The saved image data is then sent to the server.

[0471] Step 2: Receiving image data

[0472] The server receives image data sent from the terminal. The input data is an image file taken by the terminal camera. The server receives this image data in real time and temporarily stores it in storage. This storage is used as the input data source for the next analysis process.

[0473] Step 3: Image analysis

[0474] The server analyzes the received image data. The input is the received image files, and the output is the congestion level of each car and facial feature data of people passing through the ticket gates. This analysis uses a trained machine learning algorithm (e.g., convolutional neural network, CNN). The server uses a GPU to process the images at high speed and extracts the number and density of passengers and facial feature points.

[0475] Step 4: Congestion assessment

[0476] The server quantifies the congestion level of each car based on the results of image analysis. The input is the number of passengers and density data obtained from image analysis, and the output is the congestion level of each car (e.g., "Car 1 is 95% crowded," "Car 6 is 50% crowded"). Based on this numerical data, the congestion level of each car is evaluated and information is generated.

[0477] Step 5: Information Generation

[0478] The server creates announcement data based on the generated congestion status information. The input data is the congestion assessment result, and the output is information in the form of a message to be sent to the announcement system. For example, information such as "Car 1 is very crowded. Car 6 is relatively empty" is generated.

[0479] Step 6: Announcement

[0480] The server sends the created announcement data to the announcement system in the station. The input data is the message information for the announcement, and the output is the information to be displayed on the station's speakers or displays. This allows station staff and automated voice systems to provide commuters with congestion information. Users (commuters) can select their train car based on this announcement information.

[0481] Step 7: Extract facial feature points

[0482] The server analyzes the facial image data of people passing through the ticket gate and extracts facial feature points. The input data is the facial image sent from the ticket gate system camera, and the output data is the extracted facial feature points. Here, the facial feature points are extracted using a machine learning algorithm (e.g., OpenCV).

[0483] Step 8: Face data matching

[0484] The server compares the extracted facial feature points with a database of criminal records. The input data is the facial feature points and known face data in the database, and the output data is the comparison result. If a match is confirmed as a result of the comparison, the person is judged to be suspicious.

[0485] Step 9: Generate warning data

[0486] The server generates warning data based on the matching results. The input data is the facial data matching result, and the output data is a warning message. If a suspicious person is detected, the server immediately generates warning data and sends the message to the terminal (ticket gate system or station staff terminal).

[0487] Step 10: Warning Notification

[0488] The terminals (ticket gate system and station staff terminal) receive the warning data sent from the server and display the warning message. The input data is the warning message from the server, and the output is the warning message displayed on the terminal display. Station staff can check this warning and take action immediately.

[0489] The above is a specific flow of each processing step in the system of the present invention, which makes it possible to grasp the congestion situation on commuter trains and detect suspicious individuals.

[0490] (Application example 1)

[0491] 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."

[0492] In modern manufacturing factories, robots are required to efficiently transport raw materials and products, but this is often hindered by congestion within the factory and the presence of obstacles. From a safety perspective, detecting suspicious objects and obstacles is also important, but existing systems are not able to adequately address this. There is a need for a new system that can solve these problems and achieve efficient and safe transportation.

[0493] 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.

[0494] In this invention, the server includes an image acquisition unit for collecting video data of each area, an image analysis unit for analyzing the video data to determine the congestion level of each area, an information generation unit for generating information indicating the congestion level of each area based on the congestion level, a feature detection unit for detecting suspicious objects and obstacles, and a warning unit for issuing a warning about suspicious objects and obstacles detected by the feature detection unit. This makes it possible to grasp the congestion level within the factory in real time and optimize the robot's transportation route. Furthermore, detecting suspicious objects and obstacles can improve safety.

[0495] "Image acquisition means" refers to a device such as a camera mounted on a robot or peripheral equipment, and is a means for collecting image data of an area.

[0496] The "image analysis means" is a means for analyzing collected image data to determine the degree of congestion in an area and the presence or absence of suspicious objects and obstacles. Specifically, it uses machine learning algorithms to analyze images.

[0497] The "information generating means" is a means for generating information indicating the situation of an area based on the degree of congestion obtained by the image analyzing means and the detected objects.

[0498] The "announcement means" is a means for communicating the generated congestion information to robots and operators in the factory by voice or display.

[0499] The "feature detection means" is a means for detecting suspicious objects and obstacles from the collected image data.

[0500] The "warning means" is a means for issuing a warning to notify the operator when a suspicious object or obstacle is detected.

[0501] A system for realizing the present invention is configured as follows.

[0502] First, cameras mounted on the robots and peripheral equipment collect video data of each area in real time. This is the image acquisition method. The collected video data is then sent to a server via a network.

[0503] The server receives the transmitted video data and analyzes it. For the analysis, it uses machine learning algorithms such as convolutional neural networks (CNN). This allows it to determine the degree of congestion in each area and the presence of obstacles and suspicious objects. This is the image analysis method.

[0504] Next, the server generates information indicating the congestion status and safety of each area based on the information obtained by the image analysis means. This is the information generation means.

[0505] The generated information is announced to the user (factory operator or robot). Instructions for the transport route are sent to the robot in real time via the announcement means. Specifically, instructions include "Please select a different route due to high congestion" or "Please proceed along this route."

[0506] Furthermore, the server uses a feature detection means to detect suspicious objects and obstacles. If any are detected, a warning means is activated to issue a warning to the operator. This allows for a quick response and improves factory safety.

[0507] Consider the following scenario as a specific example. For example, if a specific area in a factory is congested, images captured by a camera are sent to a server in real time. The server analyzes the images and determines that the congestion level in that area is 90% or higher, and sends an instruction to the robot to "select an alternative route." If a suspicious object is detected, an alert is immediately issued and a notification is sent to the operator.

[0508] In this way, this invention makes it possible to grasp the congestion situation in a factory in real time and optimize the robot's transportation route. It also improves safety by detecting suspicious objects and obstacles.

[0509] Examples of prompts include:

[0510] "Please tell me how to implement software that analyzes the congestion status and presence of obstacles in a factory in real time. The system uses cameras, robots, and servers, and uses Python, OpenCV, and Keras. Please explain with specific code examples."

[0511] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0512] Step 1:

[0513] The terminal (camera) collects video data from each area in the factory in real time. Specifically, the camera periodically captures images of the situation within the area and acquires the image data. The input is the video data captured by the camera, and the output is the data saved on the terminal.

[0514] Step 2:

[0515] The terminal (camera) transmits the collected video data to the server via the network. The server receives the data and prepares it for the next analysis step. The input is the video data, and the output is the data sent to the server.

[0516] Step 3:

[0517] The server preprocesses the received video data, specifically resizing and normalizing the image. The input is the received video data, and the output is the preprocessed image data.

[0518] Step 4:

[0519] The server inputs the preprocessed image data into a machine learning algorithm to analyze the level of congestion, obstacles, and the presence of suspicious objects. Here, a generative AI model such as a convolutional neural network (CNN) is used. The input is the preprocessed image data, and the output is the analysis results.

[0520] Step 5:

[0521] The server generates information based on the analysis results. Specifically, it generates information indicating the congestion status of each area and the presence of suspicious objects, and stores this information in a database. The input is the analysis results, and the output is the generated information data.

[0522] Step 6:

[0523] The server notifies the users (robots and operators) of the generated information through the announcement means. The robot is informed of the transport route and the operator of safety information. The input is the generated information data and the output is notification data.

[0524] Step 7:

[0525] The device (camera) uses the feature detection means to re-detect suspicious objects and obstacles, and if any abnormalities are found, it sends the data to the server again. The input is the newly captured video data, and the output is the re-analyzed data.

[0526] Step 8:

[0527] The server analyzes the received data again, and if an abnormality is found, it issues a warning to the operator using the warning means. The input is the reanalyzed data, and the output is a warning notification.

[0528] 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.

[0529] The present invention is a system that grasps the congestion status of commuter trains in real time, detects suspicious individuals, and recognizes user emotions to adjust announcements and displays. This system includes image acquisition means, image analysis means, information generation means, announcement means, feature detection means, comparison means, warning means, and an emotion engine.

[0530] Program processing and implementation form

[0531] 1. Understanding congestion and making announcements

[0532] The terminal (in-train camera) captures real-time images from inside each car every second. This image data is compressed immediately after capture and sent to the server.

[0533] The server receives the image data sent from the terminal and analyzes it using a trained machine learning model (such as CNN) to determine the degree of congestion in each vehicle. The analysis results are used to quantify the number and density of passengers in each vehicle.

[0534] The server generates information indicating the congestion status of each car based on the analyzed congestion level. For example, "Car 1 is very crowded" or "Car 6 is relatively empty." This information is saved in a database in real time.

[0535] The server generates text data for announcements based on the congestion information and sends it to the station's announcement system. This data is processed by a voice synthesis module, and the announcement is made. Users (commuters) can select the car they want to board based on this announcement.

[0536] 2. Criminal Detection and Warning

[0537] The terminal (ticket gate system) takes facial images of all people passing through the ticket gate and transmits them to the server in real time.

[0538] The server uses a facial recognition algorithm to compare the received facial image data with a database of criminal records. If a match is found, the person is deemed suspicious and a warning message is generated. The warning message is sent to the terminal (ticket gate system) and the station staff's terminal, where a warning message or alert is immediately displayed.

[0539] 3. User emotion recognition and announcement / display adjustment

[0540] The device (emotion engine) analyzes the commuter's facial expressions and recognizes the user's emotions. The emotion engine acquires facial expression data in real time and identifies emotions such as joy, surprise, sadness, and anger.

[0541] The server stores and analyzes the recognized emotion data. For example, if many users in a certain car show angry or dissatisfied expressions, the content of the announcement regarding the level of congestion can be adjusted as much as possible. An announcement that gives a sense of security, such as "Car 6 is relatively empty. Please use it," can be generated.

[0542] The server then sends the adjusted text data for the announcement to the station announcement system, which then makes the announcement. If necessary, the server can also provide the information to station staff terminals to prompt them to take appropriate action.

[0543] Example

[0544] For example, during the morning rush hour, car 1 may be very crowded, but car 6 may be relatively empty. In this case, image data from the device (in-train camera) is sent to the server, which determines that "car 1 is very crowded" and "car 6 is relatively empty."

[0545] Furthermore, if the device (emotion engine) analyzes the user's facial expressions and detects that many users are showing signs of dissatisfaction or stress, the server will adjust the announcement. An announcement will be made that takes into consideration emotions, such as "Car 6 is relatively empty. Please use it." Users (commuters) can enjoy a comfortable commute by hearing this announcement and heading to car 6.

[0546] Furthermore, if a suspicious person is detected when passing through a ticket gate, the server will immediately issue an alert and notify the terminal (ticket gate system) and the station staff's terminal, enabling a rapid response.

[0547] In this way, the present invention will alleviate congestion, adjust announcements based on emotion recognition, and detect suspicious individuals early, thereby significantly improving the safety and comfort of commuter trains.

[0548] The processing flow will be explained below.

[0549] Specific processing flow of the program

[0550] 1. Understanding congestion and making announcements

[0551] Step 1: Image acquisition

[0552] The terminal (in-train camera) captures real-time images from inside each car every second. The image data is compressed immediately after capture and sent to the server.

[0553] Step 2: Image reception and analysis

[0554] The server receives the image data sent from the terminal, stores it in a database, and inputs the data into the analysis module.

[0555] The server's image analysis module uses a trained machine learning model (such as CNN) to analyze image data and determine the level of congestion in each vehicle. The analysis calculates the number of passengers, their density, the number of people standing, and other factors to quantify the level of congestion.

[0556] Step 3: Generate congestion information

[0557] Based on the analysis results, the server quantifies the degree of congestion in each vehicle and classifies it into categories such as "very crowded," "moderately crowded," or "empty" according to certain criteria.

[0558] Step 4: Generate announcement data

[0559] The server generates text data for announcements based on the generated congestion information. For example, "Car 1 is very crowded. Car 6 is relatively empty."

[0560] Step 5: Making the announcement

[0561] The server sends the generated text data to the station's announcement system, where it is processed by a speech synthesis module and an announcement is made.

[0562] The user (commuter) listens to the announcement and selects the car they will board. They can choose car 6 to avoid the crowded car 1.

[0563] 2. Criminal Detection and Warning

[0564] Step 1: Image acquisition

[0565] The terminal (ticket gate system) takes facial images of all people passing through the ticket gate, and the captured image data is sent to the server in real time.

[0566] Step 2: Receiving and preprocessing face images

[0567] The server receives the facial image data sent from the ticket gate system and inputs it into a facial recognition algorithm that extracts feature points.

[0568] Step 3: Matching facial feature points

[0569] The server compares the extracted feature data with a database of criminal records. The matching algorithm calculates the degree of match between the feature points and determines a match if the degree of match exceeds a certain threshold.

[0570] Step 4: Match detection and warning generation

[0571] If a match is found, the server determines that the individual is suspicious and generates warning data, including the individual's characteristics and location.

[0572] Step 5: Sending an alert

[0573] The server sends the generated warning data to the terminal (ticket gate system) and the station staff's terminal. A warning message is displayed on the ticket gate system, and an alert is sent to the station staff's terminal.

[0574] Upon receiving the alert, station staff immediately take steps to deal with suspicious persons.

[0575] 3. User emotion recognition and announcement / display adjustment

[0576] Step 1: Acquire facial expression data

[0577] The device (emotion engine) captures the facial expressions of commuters in real time, and the captured data is sent to the emotion engine.

[0578] Step 2: Facial Expression Analysis

[0579] The server receives the facial expression data sent from the emotion engine and analyzes it using an emotion recognition algorithm, identifying emotions such as joy, surprise, sadness, and anger.

[0580] Step 3: Storing and analyzing sentiment data

[0581] The server stores the recognized emotion data and performs statistical analysis. If many users in a particular vehicle express anger or dissatisfaction, the emotion data is taken into account in addition to the congestion level.

[0582] Step 4: Adjust your announcement

[0583] The server adjusts the content of the announcement based on the emotional data, for example, "Car 6 is relatively empty. Please use it." to create a sense of security.

[0584] Step 5: Making the announcement

[0585] The server then sends the adjusted text data for the announcement to the station announcement system, which then makes the announcement. If necessary, the information is also provided to station staff's terminals, encouraging them to take appropriate action.

[0586] Users (commuters) can enjoy a more comfortable commute by listening to this announcement and following the emotionally sensitive instructions.

[0587] Example

[0588] For example, during the morning rush hour, car 1 may be very crowded, while car 6 may be relatively empty. In this case, image data from the device (in-train camera) is sent to the server, which then analyzes the data and concludes that "car 1 is very crowded" and "car 6 is relatively empty."

[0589] In addition, the device (emotion engine) analyzes the user's facial expressions, and if many users show expressions of dissatisfaction or stress, the server adjusts the announcement content and makes a reassuring announcement such as, "Car 6 is relatively empty. Please use it." Users (commuters) can hear this announcement and head to car 6 for a comfortable commute.

[0590] Furthermore, if a suspicious person is detected when passing through a ticket gate, the server will immediately issue an alert and notify the terminal (ticket gate system) and the station staff's terminal, allowing for a prompt response.

[0591] As such, the present invention is a system that can alleviate congestion, adjust announcements based on emotion recognition, and detect suspicious individuals early, thereby significantly improving the safety and comfort of commuter trains.

[0592] Example 2

[0593] 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."

[0594] There is a need for a system that can efficiently grasp the congestion situation on commuter trains, detect suspicious individuals early, recognize users' emotions, and provide a comfortable commuting environment. However, conventional technologies have difficulty satisfying all of these requirements simultaneously, and have not been able to sufficiently improve the safety and comfort of commuters.

[0595] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes: image acquisition means for collecting image data of each car; image analysis means for analyzing the image data to determine the congestion level of each car; information generation means for generating information indicating the congestion level of each car based on the congestion level; announcement means for announcing the congestion level based on the information; means for saving the congestion level information in a database in real time; feature detection means for detecting features of a person passing through a ticket gate by referring to a criminal record database; comparison means for comparing the features of the person detected by the feature detection means with data in the criminal record database; warning means for issuing a warning when a match is confirmed by the comparison means; means for transmitting the warning data to the ticket gate system and the station staff terminal; emotion analysis means for analyzing the user's facial expression and recognizing their emotion; and means for adjusting the content of the announcement based on the emotion data recognized by the emotion analysis means. This makes it possible to grasp the congestion level of commuter trains in real time and provide a comfortable commuting environment by detecting suspicious individuals and recognizing the user's emotion.

[0596] "Image acquisition means" refers to a device or function for collecting image data inside a vehicle.

[0597] The "image analysis means" refers to a device or function for analyzing collected image data and determining the degree of congestion inside the vehicle.

[0598] The "information generating means" refers to a device or function for generating information indicating the congestion status of each vehicle based on the analysis results.

[0599] The "announcement means" refers to a device or function for making announcements based on the generated congestion information.

[0600] The "means for storing in a database" refers to a device or function for storing the generated congestion status information in a database in real time.

[0601] "Characteristics detection means" refers to a device or function for detecting the characteristics of a person passing through a ticket gate.

[0602] The "matching means" is a device or function for matching the characteristics of a detected person with data in a criminal record database.

[0603] The "warning means" is a device or function for issuing a warning when a match is confirmed based on the comparison result.

[0604] The "means for transmitting to the ticket gate system and the station staff terminal" refers to a device or function for transmitting warning data to the ticket gate system and the station staff terminal.

[0605] The "emotion analysis means" is a device or function for analyzing the user's facial expression and recognizing the emotion.

[0606] The "means for adjusting the content of the announcement" refers to a device or function for adjusting the content of the announcement based on the recognized emotion data.

[0607] MODE FOR CARRYING OUT THE INVENTION

[0608] The present invention is a system that grasps the congestion status of commuter trains in real time, detects suspicious individuals, and recognizes user emotions to adjust announcements and displays. This system includes image acquisition means, image analysis means, information generation means, announcement means, feature detection means, comparison means, warning means, and an emotion engine.

[0609] 1. Understanding congestion and making announcements

[0610] The device (in-train camera) captures real-time video of each car every second. This image data is compressed and sent to the server. The server then analyzes this image data using a trained machine learning model (e.g., a convolutional neural network (CNN)) to determine the level of congestion in each car. The analysis results are then quantified as the number and density of passengers in each car. The server then uses this information to generate information indicating the congestion level of each car. For example, "Car 1 is very crowded" or "Car 6 is relatively empty." This information is then stored in a database in real time. The server then generates text data for announcements based on the congestion level information and sends it to the station's announcement system. This text data is processed by a speech synthesis module and announced in real time. Users (commuters) can use this information to choose a more comfortable car.

[0611] 2. Suspicious person detection and warning

[0612] The terminal (ticket gate system) takes a facial image of every person passing through the ticket gate and sends it to a server in real time. The server then uses a facial recognition algorithm (for example, dlib, an open-source facial recognition library) to compare the received facial image data with a database of criminal records. If the comparison results in a match with a suspicious individual, the server generates warning data. This warning data is sent to the terminal (ticket gate system) and the station staff's terminal, where a warning message or alert is immediately displayed.

[0613] 3. User emotion recognition and announcement / display adjustment

[0614] The terminal (emotion engine) analyzes commuters' facial expressions and recognizes their emotions. The emotion engine uses open-source facial expression recognition libraries (e.g., OpenCV and Dlib) to acquire facial expression data in real time and identify emotions such as joy, surprise, sadness, and anger. The server stores and analyzes the recognized emotion data. For example, if many users in a certain car show angry or dissatisfied expressions, the announcement content can be adjusted. Specifically, it can be changed to a more reassuring message such as, "Car 6 is relatively empty. Please use it." This adjusted text data for the announcement is also processed by the speech synthesis module and sent to the station announcement system. Furthermore, the server can provide information to station staff terminals as needed to prompt appropriate action.

[0615] Specific examples

[0616] For example, if car 1 is very crowded during the morning rush hour, image data captured by the terminal (in-train camera) is sent to the server. The server analyzes this data and determines that "car 1 is very crowded" or "car 6 is relatively empty." Additionally, if the terminal (emotion engine) analyzes the user's facial expressions and detects that many users are expressing dissatisfaction or stress, the server adjusts the content of the announcement. A reassuring announcement such as "car 6 is relatively empty. Please use it" is made. The user (commuter) can then refer to this announcement and head to car 6, enjoying a comfortable commute. Furthermore, if a suspicious person is detected at the ticket gate, the server immediately generates warning data and sends it to the terminal (ticket gate system) and the station staff's terminal. Thus, the present invention is a system that provides a safe and comfortable commuting environment through commuter train congestion status, emotion recognition, and suspicious person detection.

[0617] Prompt Sentence Examples

[0618] "Please explain the specific processing steps of a system that monitors the congestion status of each vehicle in real time during the morning rush hour, detects suspicious individuals, and reflects user emotions in announcements."

[0619] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0620] Step 1:

[0621] The terminal (train camera) captures real-time images from inside each car every second. The input is raw video data from inside each car, and the output is compressed image data. Specifically, the camera captures images from inside each car at high resolution and compresses the data in JPEG format or similar. The compressed image data is sent to the server.

[0622] Step 2:

[0623] The server receives image data sent from the device. The input is compressed image data, and the output is data ready for analysis. The server inputs the received image data into a trained machine learning model (e.g., a convolutional neural network (CNN)) to analyze the congestion level of each vehicle. Specifically, the server uses the CNN to detect people in the image and calculate the number of people.

[0624] Step 3:

[0625] The server quantifies the degree of congestion in each car based on the image analysis results. The input is the analyzed person detection data, and the output is numerical data indicating the degree of congestion in each car. The server generates specific congestion status information, such as "Car 1 is very crowded" or "Car 6 is relatively empty." The generated information is saved in a database in real time.

[0626] Step 4:

[0627] The server creates text data for announcements based on the generated congestion information. The input is congestion information, and the output is text data for announcements. This text data is then processed by a voice synthesis module and sent to the station announcement system. Specifically, announcements such as "Car 1 is very crowded" and "Car 6 is relatively empty" are made in real time.

[0628] Step 5:

[0629] The terminal (ticket gate system) takes facial images of all people passing through the ticket gate and sends them to the server in real time. The input is the facial image of the person passing through, and the output is the transmitted facial image data. Specifically, the ticket gate system uses a high-resolution camera to take facial images, compresses the data, and sends it to the server.

[0630] Step 6:

[0631] The server analyzes the received facial image data using a facial recognition algorithm and compares it with a database of criminal records. The input is the transmitted facial image data, and the output is the comparison result. If it matches a suspicious person, the server generates warning data and sends it to the ticket gate system and station staff terminals. Specifically, the server uses the dlib library to perform facial recognition and compares it with a database of criminal records.

[0632] Step 7:

[0633] The device (emotion engine) analyzes commuters' facial expressions and recognizes their emotions. The input is facial expression data from each vehicle, and the output is analyzed emotion data. The emotion engine uses open-source facial expression recognition libraries (e.g., OpenCV and Dlib). Specifically, the device acquires facial expression data in real time and identifies emotions such as joy, surprise, sadness, and anger.

[0634] Step 8:

[0635] The server stores and analyzes the recognized emotion data. The input is the analyzed emotion data, and the output is data for adjusting the content of announcements. For example, if many users in a certain car show expressions of anger or dissatisfaction, the server adjusts the announcement content to something more reassuring, such as "Car 6 is relatively empty. Please use it." The text data for the announcement is processed by a speech synthesis module and immediately sent to the station announcement system.

[0636] Specific examples

[0637] For example, if car 1 is extremely crowded during the morning rush hour, image data from the device (in-train camera) is sent to the server, which determines that "car 1 is extremely crowded" and "car 6 is relatively empty." Furthermore, the device (emotion engine) analyzes the user's facial expressions and detects that many users are showing signs of dissatisfaction or stress. The server then adjusts the announcement to "car 6 is relatively empty. Please use it." Based on this announcement, users can head to car 6 and enjoy a comfortable commute. Furthermore, if a suspicious person is detected while passing through the ticket gate, the server immediately issues an alert and notifies the terminal (ticket gate system) and the station staff's terminal. In this way, the system grasps the situation from multiple angles in real time and responds promptly.

[0638] (Application example 2)

[0639] 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."

[0640] Conventional commuter train systems lack the means to grasp congestion status in real time and provide appropriate announcements to passengers. Furthermore, there is a need for security measures such as adjusting announcements to reduce passenger stress and anxiety, and detecting suspicious individuals. This will improve passenger comfort and safety.

[0641] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes image acquisition means for collecting image data of each vehicle, image analysis means for analyzing the image data to determine the congestion level of each vehicle, information generation means for generating information indicating the congestion level of each vehicle based on the congestion level, announcement means for announcing the congestion level based on the information, emotion recognition means for recognizing emotions by analyzing facial expressions of passengers, and announcement adjustment means for adjusting the content of announcements based on the emotions recognized by the emotion recognition means. This makes it possible to grasp the congestion level in real time, detect suspicious individuals early, and adjust announcements taking passenger emotions into consideration.

[0642] The "image acquisition means" is a mechanism that collects image data inside the vehicle using sensors such as cameras installed in each vehicle.

[0643] The "image analysis means" is a device or software for analyzing collected image data and determining the degree of congestion in each vehicle.

[0644] The "information generating means" is a mechanism that generates information indicating the congestion state of each vehicle based on the analyzed congestion degree.

[0645] The "announcement means" is a system that announces the congestion situation to passengers based on the generated congestion situation information.

[0646] The "emotion recognition means" is a device or software that analyzes the facial expressions of passengers captured by cameras inside the vehicle and recognizes their emotions.

[0647] The "announcement adjustment means" is a system that adjusts the content and manner of announcements based on the recognized emotions.

[0648] In this invention, in order to improve the comfort and safety of self-driving vehicles, a system is constructed that grasps the congestion situation in each vehicle and the emotions of passengers in real time, and makes appropriate announcements and displays based on that information.

[0649] The server includes the following means:

[0650] 1. Image acquisition method:

[0651] Real-time images are acquired using cameras installed inside the vehicle.

[0652] Cameras are installed at various locations within the vehicle to provide comprehensive coverage of the situation within each vehicle.

[0653] 2. Image analysis methods:

[0654] The acquired image data is compressed and transmitted to the server.

[0655] The server analyzes the image data using machine learning algorithms, such as convolutional neural networks (CNNs), to determine the level of congestion in each vehicle.

[0656] 3. Information generation means:

[0657] Based on the analysis results, information indicating the congestion status of each car is generated, such as "Car 1 is very crowded" or "Car 6 is relatively empty."

[0658] The generated information is stored in a database on the server in real time.

[0659] 4. Announcement Method:

[0660] Text data for announcements is generated based on the information indicating the congestion state.

[0661] The generated text data is converted into speech via a speech synthesis module and transmitted to the announcement system inside the train or station.

[0662] 5. Emotion recognition means:

[0663] The server uses footage from cameras inside the vehicle to analyze passengers' facial expressions in real time and recognize emotions such as joy, anger, sadness, and surprise.

[0664] Emotion recognition uses techniques such as convolutional neural networks (CNNs) for emotion classification.

[0665] 6. Announcement Coordination Methods:

[0666] Adjust the content and manner of announcements based on the recognized emotional data.

[0667] For example, if many passengers look dissatisfied, an announcement will be made saying, "The next car is available, so please use it."

[0668] As a concrete example, imagine a situation where car 1 is very crowded during the morning rush hour, but car 6 is relatively empty. Image data captured by the camera is analyzed by the server, and information such as "car 1 is very crowded" and "car 6 is relatively empty" is generated. This information is provided to passengers as a voice announcement. Furthermore, if the passenger's facial expression indicates dissatisfaction, the server will make a more considerate announcement.

[0669] An example of a prompt is shown below.

[0670] "We will develop a system that analyzes the congestion situation and passenger emotions in real time from camera footage of autonomous vehicles, and adjusts the content of announcements and displays based on that analysis. For example, if the congestion level is high and passengers are expressing dissatisfaction, the system will announce that the next car is available, improving passenger comfort."

[0671] In this way, it is possible to realize a system that enhances the comfort and safety of autonomous vehicles and improves passenger satisfaction.

[0672] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0673] Step 1:

[0674] The terminal (in-car camera) captures real-time images from inside the car. The camera is installed in a specific position inside each car and captures the overall congestion situation and the facial expressions of passengers. The camera captures image data every second, compresses this data, and sends it to the server.

[0675] Input: Real-time video from inside the vehicle

[0676] Output: Compressed image data

[0677] Step 2:

[0678] The server receives the image data sent from the device. Each received image is analyzed using a machine learning algorithm. Specifically, a convolutional neural network (CNN) model is used to determine the degree of vehicle congestion.

[0679] Input: Compressed image data

[0680] Action: Data extraction and image analysis

[0681] Output: Quantified congestion data

[0682] Step 3:

[0683] The server generates information indicating the congestion status of each car based on the analysis results. For example, if the analysis shows that car 1 is highly crowded, it generates information such as "car 1 is very crowded."

[0684] Input: Data quantifying congestion

[0685] Action: Generate congestion information

[0686] Output: Information showing the occupancy status of each vehicle

[0687] Step 4:

[0688] The server then generates text data for announcements based on the congestion status, which is then converted into speech by a speech synthesis module and sent to the announcement system.

[0689] Input: Information indicating the occupancy status of each vehicle

[0690] Operation: Generation of text data for announcements and speech synthesis

[0691] Output: Announcement audio data

[0692] Step 5:

[0693] The terminals (speakers inside the train and in stations) receive the announcement audio data sent from the server and make the actual announcements, allowing passengers to understand the congestion situation in real time.

[0694] Input: Announcement audio data

[0695] Action: Make an announcement

[0696] Output: Announcement voice

[0697] Step 6:

[0698] The terminal (in-vehicle camera) captures facial expression data of passengers. Similarly, image data is captured every second, compressed, and sent to the server.

[0699] Input: Real-time video of passenger facial expressions

[0700] Output: Compressed facial expression data

[0701] Step 7:

[0702] The server receives the facial expression data and uses a machine learning model for emotion recognition (e.g., emotion classification model) to recognize emotions such as joy, anger, sadness, and surprise.

[0703] Input: Compressed facial expression data

[0704] Action: Facial expression analysis and emotion recognition

[0705] Output: Emotion data for each passenger

[0706] Step 8:

[0707] The server then adjusts the announcement content based on the recognized emotion data. For example, if many passengers show expressions of dissatisfaction or stress, the announcement will be changed to one that promotes a sense of security, such as "The next car is available, so please use it."

[0708] Input: Emotion data for each passenger

[0709] Action: Adjust announcement content

[0710] Output: Adjusted text data for announcement

[0711] Step 9:

[0712] The server converts the adjusted text data for announcement into voice again using the voice synthesis module and transmits it to the announcement system.

[0713] Input: Adjusted text data for announcement

[0714] Operation: Generation of announcement voice data and speech synthesis

[0715] Output: Adjusted announcement audio data

[0716] Step 10:

[0717] The terminals (speakers inside the train and at stations) receive the adjusted announcement audio data sent from the server and make announcements based on it.

[0718] Input: Adjusted announcement audio data

[0719] Action: Announcement is performed after adjustment

[0720] Output: Adjusted announcement audio

[0721] In this way, a system is built that can grasp congestion status in real time and make appropriate announcements based on that information. In addition, by adjusting the content of announcements to take passengers' emotions into consideration, a comfortable riding environment is provided.

[0722] 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.

[0723] 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.

[0724] 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.

[0725] [Third embodiment]

[0726] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0727] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0728] 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).

[0729] 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.

[0730] 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.

[0731] 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).

[0732] 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.

[0733] 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.

[0734] 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.

[0735] 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.

[0736] 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.

[0737] 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."

[0738] The present invention provides a system for grasping the congestion status of commuter trains in real time and detecting suspicious individuals. The system includes an image acquisition unit, an image analysis unit, an information generation unit, an announcement unit, a feature detection unit, a comparison unit, and a warning unit.

[0739] Program processing and implementation form

[0740] 1. Image acquisition method

[0741] The terminal (in-train camera) periodically captures image data of each car. For example, a camera installed in each car captures an image of passengers every second and collects this image data.

[0742] 2. Image analysis methods

[0743] The server analyzes the image data sent from the device in real time using a trained machine learning model (e.g., convolutional neural network, CNN).The server uses the image data to determine the degree of congestion in each vehicle and quantifies the number and density of passengers in each vehicle.

[0744] 3. Information generation means

[0745] Based on the analyzed congestion level, the server generates information indicating the congestion status of each car (e.g., "Car 1 is 95% crowded," "Car 6 is 50% crowded," etc.). This information is stored in a database in real time.

[0746] 4. Announcement Methods

[0747] The server sends the generated congestion status information to the station announcement system. This allows station staff or automated announcements to provide users with information such as "Car 1 is very crowded. Car 6 is relatively empty." Users (commuters) can then select the car they want to board based on the announcements.

[0748] 5. Feature detection methods

[0749] The terminal (ticket gate system) captures a facial image of a person passing through the ticket gate, and the captured facial image data is sent to a server in real time.

[0750] 6. Verification Method

[0751] The server uses a facial recognition algorithm to extract features from the received facial image and compares it with a database of criminal records, which contains facial feature data of past criminals.

[0752] 7. Warning measures

[0753] If the server finds a match, it determines the person is suspicious and generates a warning. A warning message is displayed on the terminal (ticket gate system), and a warning notification is simultaneously sent to the station staff's terminal. This alert allows station staff to respond immediately and prevent crimes from occurring.

[0754] Example

[0755] For example, during rush hour on a certain weekday morning, car 1 is very crowded, but car 6 is relatively empty. In this case, image data taken by the device is sent to the server, which analyzes it and determines that "car 1 is 95% crowded" and "car 6 is 50% crowded." Based on this, an announcement is made that "car 6 is relatively empty." The user (commuter) can select car 6 based on this information to enjoy a comfortable commute.

[0756] Furthermore, if a suspicious person is detected while passing through a ticket gate, the server immediately sends an alert, which is displayed on the terminal (ticket gate system) and the station staff's terminal.Station staff can then respond quickly and prevent crimes from occurring.

[0757] Thus, the present invention is an effective means for alleviating congestion on commuter trains and improving safety.

[0758] The processing flow will be explained below.

[0759] Specific processing flow of the program

[0760] 1. Understanding congestion and making announcements

[0761] Step 1: Image acquisition

[0762] The terminal (in-train camera) captures real-time images from inside each car every second. This image data is compressed immediately after capture and sent to the server.

[0763] Step 2: Image reception and analysis

[0764] The server receives the image data sent from the terminal, immediately stores it in a database, and inputs it into the analysis module.

[0765] The server's image analysis module uses machine learning models (such as CNN) to analyze the congestion level of each vehicle, calculating the number of passengers, their density, the number of people standing, and other factors to quantify the level of congestion.

[0766] Step 3: Generate congestion information

[0767] Based on the analysis results, the server quantifies the degree of congestion in each vehicle and classifies it into categories such as "very crowded," "moderately crowded," or "empty" according to certain criteria.

[0768] Step 4: Generate announcement data

[0769] The server generates text data for announcements based on the generated congestion information. For example, "Car 1 is very crowded. Car 6 is relatively empty."

[0770] Step 5: Making the announcement

[0771] The server sends the generated text data to the station's announcement system, where it is processed by a speech synthesis module and an announcement is made.

[0772] Users (commuters) listen to the announcement and select the car they want to get in. For example, if they know that car 1 is crowded, they can choose car 6.

[0773] 2. Criminal Detection and Warning

[0774] Step 1: Image acquisition

[0775] The terminal (ticket gate system) takes facial images of all people passing through the ticket gate, and the captured image data is sent to the server in real time.

[0776] Step 2: Receiving and preprocessing face images

[0777] The server receives the facial image data sent from the ticket gate system, and immediately extracts facial feature points using a feature point extraction algorithm.

[0778] Step 3: Matching facial feature points

[0779] The server compares the minutiae data with a criminal record database. A matching algorithm calculates the degree of match between the minutiae and determines a match if the degree of match exceeds a certain threshold.

[0780] Step 4: Match detection and warning generation

[0781] If a match is found, the server generates alert data, which includes the suspect's characteristics and location.

[0782] Step 5: Sending an alert

[0783] The server sends the generated warning data to the terminal (ticket gate system) and the station staff's terminal. A warning message is displayed on the ticket gate system, and an alert is sent to the station staff's terminal.

[0784] Upon receiving the alert, station staff immediately take steps to deal with suspicious persons.

[0785] The above is the specific operation of each step. This system makes full use of image analysis and facial recognition technology to reduce congestion on commuter trains and improve safety.

[0786] Example 1

[0787] 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."

[0788] On current commuter trains, it is difficult to grasp the congestion status of each car in real time, which results in passengers often boarding cars that are more crowded. Also, if a suspicious person passes through the ticket gate, there are limited ways to quickly deal with the problem, which poses safety issues.

[0789] 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.

[0790] In this invention, the server includes image acquisition means for collecting image data of each vehicle, image analysis means for analyzing the image data to determine the degree of congestion of each vehicle, information generation means for generating information indicating the congestion status of each vehicle based on the congestion status, announcement means for announcing the congestion status based on the information, image acquisition means for taking facial images of people passing through the ticket gate, comparison means for analyzing the facial images to extract feature points and comparing them with a criminal record database, and warning means for generating and notifying warning data if a match is confirmed by the comparison means. This allows passengers to choose vehicles that are less crowded, improving safety.

[0791] The "image acquisition means" is a device for capturing and collecting image data of passengers and passersby inside trains and at ticket gates.

[0792] "Image analysis means" is a technology for analyzing acquired image data and extracting specific information (for example, congestion level or characteristic points).

[0793] An "information generation means" is a process or system for generating useful information (e.g., congestion status) from analyzed data.

[0794] "Announcement means" refers to audio and display devices used to inform passengers and other relevant parties of the situation based on the generated information.

[0795] The "matching means" is a system for comparing the feature points of the acquired facial image with an existing database to confirm a match.

[0796] The "warning means" is a mechanism for generating warning data and notifying relevant parties when a match is confirmed.

[0797] "Crowding level of each car" is an index that indicates the number and density of passengers in each car of a train.

[0798] A "criminal record database" is a database that registers facial images and feature points of people who have a criminal record.

[0799] "Feature detection means" is a technology for analyzing and extracting feature points from the facial image of a person passing through a ticket gate.

[0800] This invention is a system for understanding the congestion situation of commuter trains in real time and detecting suspicious individuals. This system operates through a series of processes: image data acquisition, analysis, and warning notification. Each function is explained in detail below.

[0801] Image Acquisition Method

[0802] The devices (in-train cameras and ticket gate system cameras) periodically capture image data of passengers and passersby and send it to a server. For example, high-resolution cameras (generic name) installed in each car and ticket gate system cameras (generic name) are used. These cameras automatically use a timer function to periodically capture images, store the data in temporary storage, and then send it to the server.

[0803] Image analysis methods

[0804] The server analyzes the received image data in real time. This analysis uses a pre-trained machine learning algorithm (e.g., convolutional neural network, generic name). Software used for the analysis includes TensorFlow (generic name). The received image data is processed by a server equipped with a high-performance GPU (generic name), which quantifies the number of passengers and their density to determine the level of congestion.

[0805] Information generation means

[0806] The server generates information showing the congestion status of each car based on the analyzed congestion level information. For example, information such as "Car 1 is 95% crowded" or "Car 6 is 50% crowded" is generated. This information is saved in a database in real time and kept up to date. MySQL (generic name) or similar is used as the database.

[0807] Announcement Method

[0808] The server sends the generated congestion status information to the announcement system within the station. This allows station staff or automated voice systems to provide users with information such as, "Car 1 is very crowded. Car 6 is relatively empty." Users (passengers) can choose their own car based on this information, allowing them to avoid crowds and enjoy a comfortable commute.

[0809] Feature detection and matching methods

[0810] The terminal (ticket gate system) takes a picture of the face of a person passing through the ticket gate and sends it to the server. The server analyzes the received face image using a facial recognition algorithm (e.g., OpenCV, generic name) and extracts feature points. The extracted feature points are then compared with a database of criminal records.

[0811] warning means

[0812] If the server finds a match, it determines the person is suspicious and generates a warning. The generated warning data is sent in real time to the ticket gate system and station staff terminals, where a warning message is displayed. This allows station staff to respond quickly and prevent crimes from occurring.

[0813] Specific examples

[0814] For example, during rush hour on a certain weekday morning, car 1 is very crowded, but car 6 is relatively empty. In this case, image data taken by the device is sent to the server, which analyzes it and determines that "car 1 is 95% crowded" and "car 6 is 50% crowded." Based on this, an announcement is made that "car 6 is relatively empty." The user (commuter) can select car 6 based on this information to enjoy a comfortable commute.

[0815] Furthermore, if a suspicious person is detected while passing through a ticket gate, the server immediately sends an alert, which is displayed on the ticket gate system and on the station staff's terminals, allowing station staff to respond quickly and prevent crimes from occurring.

[0816] Prompt Sentence Examples

[0817] "Please explain the system that monitors the congestion status of train cars in real time during the weekday morning rush hour and selects the appropriate car. Please also explain the process for detecting suspicious individuals at ticket gates and issuing warnings."

[0818] Thus, the present invention is a very effective means for alleviating congestion on commuter trains and improving safety.

[0819] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0820] Step 1: Image acquisition

[0821] The terminals (in-train cameras and ticket gate system cameras) capture image data of each car and people passing through the ticket gates. The input data are still images taken by the cameras, which are saved in temporary storage. Specifically, the cameras record images every second and temporarily save these images in memory. The saved image data is then sent to the server.

[0822] Step 2: Receiving image data

[0823] The server receives image data sent from the terminal. The input data is an image file taken by the terminal camera. The server receives this image data in real time and temporarily stores it in storage. This storage is used as the input data source for the next analysis process.

[0824] Step 3: Image analysis

[0825] The server analyzes the received image data. The input is the received image files, and the output is the congestion level of each car and facial feature data of people passing through the ticket gates. This analysis uses a trained machine learning algorithm (e.g., convolutional neural network, CNN). The server uses a GPU to process the images at high speed and extracts the number and density of passengers and facial feature points.

[0826] Step 4: Congestion assessment

[0827] The server quantifies the congestion level of each car based on the results of image analysis. The input is the number of passengers and density data obtained from image analysis, and the output is the congestion level of each car (e.g., "Car 1 is 95% crowded," "Car 6 is 50% crowded"). Based on this numerical data, the congestion level of each car is evaluated and information is generated.

[0828] Step 5: Information Generation

[0829] The server creates announcement data based on the generated congestion status information. The input data is the congestion assessment result, and the output is information in the form of a message to be sent to the announcement system. For example, information such as "Car 1 is very crowded. Car 6 is relatively empty" is generated.

[0830] Step 6: Announcement

[0831] The server sends the created announcement data to the announcement system in the station. The input data is the message information for the announcement, and the output is the information to be displayed on the station's speakers or displays. This allows station staff and automated voice systems to provide commuters with congestion information. Users (commuters) can select their train car based on this announcement information.

[0832] Step 7: Extract facial feature points

[0833] The server analyzes the facial image data of people passing through the ticket gate and extracts facial feature points. The input data is the facial image sent from the ticket gate system camera, and the output data is the extracted facial feature points. Here, the facial feature points are extracted using a machine learning algorithm (e.g., OpenCV).

[0834] Step 8: Face data matching

[0835] The server compares the extracted facial feature points with a database of criminal records. The input data is the facial feature points and known face data in the database, and the output data is the comparison result. If a match is confirmed as a result of the comparison, the person is judged to be suspicious.

[0836] Step 9: Generate warning data

[0837] The server generates warning data based on the matching results. The input data is the facial data matching result, and the output data is a warning message. If a suspicious person is detected, the server immediately generates warning data and sends the message to the terminal (ticket gate system or station staff terminal).

[0838] Step 10: Warning Notification

[0839] The terminals (ticket gate system and station staff terminal) receive the warning data sent from the server and display the warning message. The input data is the warning message from the server, and the output is the warning message displayed on the terminal display. Station staff can check this warning and take action immediately.

[0840] The above is a specific flow of each processing step in the system of the present invention, which makes it possible to grasp the congestion situation on commuter trains and detect suspicious individuals.

[0841] (Application example 1)

[0842] 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."

[0843] In modern manufacturing factories, robots are required to efficiently transport raw materials and products, but this is often hindered by congestion within the factory and the presence of obstacles. From a safety perspective, detecting suspicious objects and obstacles is also important, but existing systems are not able to adequately address this. There is a need for a new system that can solve these problems and achieve efficient and safe transportation.

[0844] 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.

[0845] In this invention, the server includes an image acquisition unit for collecting video data of each area, an image analysis unit for analyzing the video data to determine the congestion level of each area, an information generation unit for generating information indicating the congestion level of each area based on the congestion level, a feature detection unit for detecting suspicious objects and obstacles, and a warning unit for issuing a warning about suspicious objects and obstacles detected by the feature detection unit. This makes it possible to grasp the congestion level within the factory in real time and optimize the robot's transportation route. Furthermore, detecting suspicious objects and obstacles can improve safety.

[0846] "Image acquisition means" refers to a device such as a camera mounted on a robot or peripheral equipment, and is a means for collecting image data of an area.

[0847] The "image analysis means" is a means for analyzing collected image data to determine the degree of congestion in an area and the presence or absence of suspicious objects and obstacles. Specifically, it uses machine learning algorithms to analyze images.

[0848] The "information generating means" is a means for generating information indicating the situation of an area based on the degree of congestion obtained by the image analyzing means and the detected objects.

[0849] The "announcement means" is a means for communicating the generated congestion information to robots and operators in the factory by voice or display.

[0850] The "feature detection means" is a means for detecting suspicious objects and obstacles from the collected image data.

[0851] The "warning means" is a means for issuing a warning to notify the operator when a suspicious object or obstacle is detected.

[0852] A system for realizing the present invention is configured as follows.

[0853] First, cameras mounted on the robots and peripheral equipment collect video data of each area in real time. This is the image acquisition method. The collected video data is then sent to a server via a network.

[0854] The server receives the transmitted video data and analyzes it. For the analysis, it uses machine learning algorithms such as convolutional neural networks (CNN). This allows it to determine the degree of congestion in each area and the presence of obstacles and suspicious objects. This is the image analysis method.

[0855] Next, the server generates information indicating the congestion status and safety of each area based on the information obtained by the image analysis means. This is the information generation means.

[0856] The generated information is announced to the user (factory operator or robot). Instructions for the transport route are sent to the robot in real time via the announcement means. Specifically, instructions include "Please select a different route due to high congestion" or "Please proceed along this route."

[0857] Furthermore, the server uses a feature detection means to detect suspicious objects and obstacles. If any are detected, a warning means is activated to issue a warning to the operator. This allows for a quick response and improves factory safety.

[0858] Consider the following scenario as a specific example. For example, if a specific area in a factory is congested, images captured by a camera are sent to a server in real time. The server analyzes the images and determines that the congestion level in that area is 90% or higher, and sends an instruction to the robot to "select an alternative route." If a suspicious object is detected, an alert is immediately issued and a notification is sent to the operator.

[0859] In this way, this invention makes it possible to grasp the congestion situation in a factory in real time and optimize the robot's transportation route. It also improves safety by detecting suspicious objects and obstacles.

[0860] Examples of prompts include:

[0861] "Please tell me how to implement software that analyzes the congestion status and presence of obstacles in a factory in real time. The system uses cameras, robots, and servers, and uses Python, OpenCV, and Keras. Please explain with specific code examples."

[0862] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0863] Step 1:

[0864] The terminal (camera) collects video data from each area in the factory in real time. Specifically, the camera periodically captures images of the situation within the area and acquires the image data. The input is the video data captured by the camera, and the output is the data saved on the terminal.

[0865] Step 2:

[0866] The terminal (camera) transmits the collected video data to the server via the network. The server receives the data and prepares it for the next analysis step. The input is the video data, and the output is the data sent to the server.

[0867] Step 3:

[0868] The server preprocesses the received video data, specifically resizing and normalizing the image. The input is the received video data, and the output is the preprocessed image data.

[0869] Step 4:

[0870] The server inputs the preprocessed image data into a machine learning algorithm to analyze the level of congestion, obstacles, and the presence of suspicious objects. Here, a generative AI model such as a convolutional neural network (CNN) is used. The input is the preprocessed image data, and the output is the analysis results.

[0871] Step 5:

[0872] The server generates information based on the analysis results. Specifically, it generates information indicating the congestion status of each area and the presence of suspicious objects, and stores this information in a database. The input is the analysis results, and the output is the generated information data.

[0873] Step 6:

[0874] The server notifies the users (robots and operators) of the generated information through the announcement means. The robot is informed of the transport route and the operator of safety information. The input is the generated information data and the output is notification data.

[0875] Step 7:

[0876] The device (camera) uses the feature detection means to re-detect suspicious objects and obstacles, and if any abnormalities are found, it sends the data to the server again. The input is the newly captured video data, and the output is the re-analyzed data.

[0877] Step 8:

[0878] The server analyzes the received data again, and if an abnormality is found, it issues a warning to the operator using the warning means. The input is the reanalyzed data, and the output is a warning notification.

[0879] 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.

[0880] The present invention is a system that grasps the congestion status of commuter trains in real time, detects suspicious individuals, and recognizes user emotions to adjust announcements and displays. This system includes image acquisition means, image analysis means, information generation means, announcement means, feature detection means, comparison means, warning means, and an emotion engine.

[0881] Program processing and implementation form

[0882] 1. Understanding congestion and making announcements

[0883] The terminal (in-train camera) captures real-time images from inside each car every second. This image data is compressed immediately after capture and sent to the server.

[0884] The server receives the image data sent from the terminal and analyzes it using a trained machine learning model (such as CNN) to determine the degree of congestion in each vehicle. The analysis results are used to quantify the number and density of passengers in each vehicle.

[0885] The server generates information indicating the congestion status of each car based on the analyzed congestion level. For example, "Car 1 is very crowded" or "Car 6 is relatively empty." This information is saved in a database in real time.

[0886] The server generates text data for announcements based on the congestion information and sends it to the station's announcement system. This data is processed by a voice synthesis module, and the announcement is made. Users (commuters) can select the car they want to board based on this announcement.

[0887] 2. Criminal Detection and Warning

[0888] The terminal (ticket gate system) takes facial images of all people passing through the ticket gate and transmits them to the server in real time.

[0889] The server uses a facial recognition algorithm to compare the received facial image data with a database of criminal records. If a match is found, the person is deemed suspicious and a warning message is generated. The warning message is sent to the terminal (ticket gate system) and the station staff's terminal, where a warning message or alert is immediately displayed.

[0890] 3. User emotion recognition and announcement / display adjustment

[0891] The device (emotion engine) analyzes the commuter's facial expressions and recognizes the user's emotions. The emotion engine acquires facial expression data in real time and identifies emotions such as joy, surprise, sadness, and anger.

[0892] The server stores and analyzes the recognized emotion data. For example, if many users in a certain car show angry or dissatisfied expressions, the content of the announcement regarding the level of congestion can be adjusted as much as possible. An announcement that gives a sense of security, such as "Car 6 is relatively empty. Please use it," can be generated.

[0893] The server then sends the adjusted text data for the announcement to the station announcement system, which then makes the announcement. If necessary, the server can also provide the information to station staff terminals to prompt them to take appropriate action.

[0894] Example

[0895] For example, during the morning rush hour, car 1 may be very crowded, but car 6 may be relatively empty. In this case, image data from the device (in-train camera) is sent to the server, which determines that "car 1 is very crowded" and "car 6 is relatively empty."

[0896] Furthermore, if the device (emotion engine) analyzes the user's facial expressions and detects that many users are showing signs of dissatisfaction or stress, the server will adjust the announcement. An announcement will be made that takes into consideration emotions, such as "Car 6 is relatively empty. Please use it." Users (commuters) can enjoy a comfortable commute by hearing this announcement and heading to car 6.

[0897] Furthermore, if a suspicious person is detected when passing through a ticket gate, the server will immediately issue an alert and notify the terminal (ticket gate system) and the station staff's terminal, enabling a rapid response.

[0898] In this way, the present invention will alleviate congestion, adjust announcements based on emotion recognition, and detect suspicious individuals early, thereby significantly improving the safety and comfort of commuter trains.

[0899] The processing flow will be explained below.

[0900] Specific processing flow of the program

[0901] 1. Understanding congestion and making announcements

[0902] Step 1: Image acquisition

[0903] The terminal (in-train camera) captures real-time images from inside each car every second. The image data is compressed immediately after capture and sent to the server.

[0904] Step 2: Image reception and analysis

[0905] The server receives the image data sent from the terminal, stores it in a database, and inputs the data into the analysis module.

[0906] The server's image analysis module uses a trained machine learning model (such as CNN) to analyze image data and determine the level of congestion in each vehicle. The analysis calculates the number of passengers, their density, the number of people standing, and other factors to quantify the level of congestion.

[0907] Step 3: Generate congestion information

[0908] Based on the analysis results, the server quantifies the degree of congestion in each vehicle and classifies it into categories such as "very crowded," "moderately crowded," or "empty" according to certain criteria.

[0909] Step 4: Generate announcement data

[0910] The server generates text data for announcements based on the generated congestion information. For example, "Car 1 is very crowded. Car 6 is relatively empty."

[0911] Step 5: Making the announcement

[0912] The server sends the generated text data to the station's announcement system, where it is processed by a speech synthesis module and an announcement is made.

[0913] The user (commuter) listens to the announcement and selects the car they will board. They can choose car 6 to avoid the crowded car 1.

[0914] 2. Criminal Detection and Warning

[0915] Step 1: Image acquisition

[0916] The terminal (ticket gate system) takes facial images of all people passing through the ticket gate, and the captured image data is sent to the server in real time.

[0917] Step 2: Receiving and preprocessing face images

[0918] The server receives the facial image data sent from the ticket gate system and inputs it into a facial recognition algorithm that extracts feature points.

[0919] Step 3: Matching facial feature points

[0920] The server compares the extracted feature data with a database of criminal records. The matching algorithm calculates the degree of match between the feature points and determines a match if the degree of match exceeds a certain threshold.

[0921] Step 4: Match detection and warning generation

[0922] If a match is found, the server determines that the individual is suspicious and generates warning data, including the individual's characteristics and location.

[0923] Step 5: Sending an alert

[0924] The server sends the generated warning data to the terminal (ticket gate system) and the station staff's terminal. A warning message is displayed on the ticket gate system, and an alert is sent to the station staff's terminal.

[0925] Upon receiving the alert, station staff immediately take steps to deal with suspicious persons.

[0926] 3. User emotion recognition and announcement / display adjustment

[0927] Step 1: Acquire facial expression data

[0928] The device (emotion engine) captures the facial expressions of commuters in real time, and the captured data is sent to the emotion engine.

[0929] Step 2: Facial Expression Analysis

[0930] The server receives the facial expression data sent from the emotion engine and analyzes it using an emotion recognition algorithm, identifying emotions such as joy, surprise, sadness, and anger.

[0931] Step 3: Storing and analyzing sentiment data

[0932] The server stores the recognized emotion data and performs statistical analysis. If many users in a particular vehicle express anger or dissatisfaction, the emotion data is taken into account in addition to the congestion level.

[0933] Step 4: Adjust your announcement

[0934] The server adjusts the content of the announcement based on the emotional data, for example, "Car 6 is relatively empty. Please use it." to create a sense of security.

[0935] Step 5: Making the announcement

[0936] The server then sends the adjusted text data for the announcement to the station announcement system, which then makes the announcement. If necessary, the information is also provided to station staff's terminals, encouraging them to take appropriate action.

[0937] Users (commuters) can enjoy a more comfortable commute by listening to this announcement and following the emotionally sensitive instructions.

[0938] Example

[0939] For example, during the morning rush hour, car 1 may be very crowded, while car 6 may be relatively empty. In this case, image data from the device (in-train camera) is sent to the server, which then analyzes the data and concludes that "car 1 is very crowded" and "car 6 is relatively empty."

[0940] In addition, the device (emotion engine) analyzes the user's facial expressions, and if many users show expressions of dissatisfaction or stress, the server adjusts the announcement content and makes a reassuring announcement such as, "Car 6 is relatively empty. Please use it." Users (commuters) can hear this announcement and head to car 6 for a comfortable commute.

[0941] Furthermore, if a suspicious person is detected when passing through a ticket gate, the server will immediately issue an alert and notify the terminal (ticket gate system) and the station staff's terminal, allowing for a prompt response.

[0942] As such, the present invention is a system that can alleviate congestion, adjust announcements based on emotion recognition, and detect suspicious individuals early, thereby significantly improving the safety and comfort of commuter trains.

[0943] Example 2

[0944] 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."

[0945] There is a need for a system that can efficiently grasp the congestion situation on commuter trains, detect suspicious individuals early, recognize users' emotions, and provide a comfortable commuting environment. However, conventional technologies have difficulty satisfying all of these requirements simultaneously, and have not been able to sufficiently improve the safety and comfort of commuters.

[0946] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes: image acquisition means for collecting image data of each car; image analysis means for analyzing the image data to determine the congestion level of each car; information generation means for generating information indicating the congestion level of each car based on the congestion level; announcement means for announcing the congestion level based on the information; means for saving the congestion level information in a database in real time; feature detection means for detecting features of a person passing through a ticket gate by referring to a criminal record database; comparison means for comparing the features of the person detected by the feature detection means with data in the criminal record database; warning means for issuing a warning when a match is confirmed by the comparison means; means for transmitting the warning data to the ticket gate system and the station staff terminal; emotion analysis means for analyzing the user's facial expression and recognizing their emotion; and means for adjusting the content of the announcement based on the emotion data recognized by the emotion analysis means. This makes it possible to grasp the congestion level of commuter trains in real time and provide a comfortable commuting environment by detecting suspicious individuals and recognizing the user's emotion.

[0947] "Image acquisition means" refers to a device or function for collecting image data inside a vehicle.

[0948] The "image analysis means" refers to a device or function for analyzing collected image data and determining the degree of congestion inside the vehicle.

[0949] The "information generating means" refers to a device or function for generating information indicating the congestion status of each vehicle based on the analysis results.

[0950] The "announcement means" refers to a device or function for making announcements based on the generated congestion information.

[0951] The "means for storing in a database" refers to a device or function for storing the generated congestion status information in a database in real time.

[0952] "Characteristics detection means" refers to a device or function for detecting the characteristics of a person passing through a ticket gate.

[0953] The "matching means" is a device or function for matching the characteristics of a detected person with data in a criminal record database.

[0954] The "warning means" is a device or function for issuing a warning when a match is confirmed based on the comparison result.

[0955] The "means for transmitting to the ticket gate system and the station staff terminal" refers to a device or function for transmitting warning data to the ticket gate system and the station staff terminal.

[0956] The "emotion analysis means" is a device or function for analyzing the user's facial expression and recognizing the emotion.

[0957] The "means for adjusting the content of the announcement" refers to a device or function for adjusting the content of the announcement based on the recognized emotion data.

[0958] MODE FOR CARRYING OUT THE INVENTION

[0959] The present invention is a system that grasps the congestion status of commuter trains in real time, detects suspicious individuals, and recognizes user emotions to adjust announcements and displays. This system includes image acquisition means, image analysis means, information generation means, announcement means, feature detection means, comparison means, warning means, and an emotion engine.

[0960] 1. Understanding congestion and making announcements

[0961] The device (in-train camera) captures real-time video of each car every second. This image data is compressed and sent to the server. The server then analyzes this image data using a trained machine learning model (e.g., a convolutional neural network (CNN)) to determine the level of congestion in each car. The analysis results are then quantified as the number and density of passengers in each car. The server then uses this information to generate information indicating the congestion level of each car. For example, "Car 1 is very crowded" or "Car 6 is relatively empty." This information is then stored in a database in real time. The server then generates text data for announcements based on the congestion level information and sends it to the station's announcement system. This text data is processed by a speech synthesis module and announced in real time. Users (commuters) can use this information to choose a more comfortable car.

[0962] 2. Suspicious person detection and warning

[0963] The terminal (ticket gate system) takes a facial image of every person passing through the ticket gate and sends it to a server in real time. The server then uses a facial recognition algorithm (for example, dlib, an open-source facial recognition library) to compare the received facial image data with a database of criminal records. If the comparison results in a match with a suspicious individual, the server generates warning data. This warning data is sent to the terminal (ticket gate system) and the station staff's terminal, where a warning message or alert is immediately displayed.

[0964] 3. User emotion recognition and announcement / display adjustment

[0965] The terminal (emotion engine) analyzes commuters' facial expressions and recognizes their emotions. The emotion engine uses open-source facial expression recognition libraries (e.g., OpenCV and Dlib) to acquire facial expression data in real time and identify emotions such as joy, surprise, sadness, and anger. The server stores and analyzes the recognized emotion data. For example, if many users in a certain car show angry or dissatisfied expressions, the announcement content can be adjusted. Specifically, it can be changed to a more reassuring message such as, "Car 6 is relatively empty. Please use it." This adjusted text data for the announcement is also processed by the speech synthesis module and sent to the station announcement system. Furthermore, the server can provide information to station staff terminals as needed to prompt appropriate action.

[0966] Specific examples

[0967] For example, if car 1 is very crowded during the morning rush hour, image data captured by the terminal (in-train camera) is sent to the server. The server analyzes this data and determines that "car 1 is very crowded" or "car 6 is relatively empty." Additionally, if the terminal (emotion engine) analyzes the user's facial expressions and detects that many users are expressing dissatisfaction or stress, the server adjusts the content of the announcement. A reassuring announcement such as "car 6 is relatively empty. Please use it" is made. The user (commuter) can then refer to this announcement and head to car 6, enjoying a comfortable commute. Furthermore, if a suspicious person is detected at the ticket gate, the server immediately generates warning data and sends it to the terminal (ticket gate system) and the station staff's terminal. Thus, the present invention is a system that provides a safe and comfortable commuting environment through commuter train congestion status, emotion recognition, and suspicious person detection.

[0968] Prompt Sentence Examples

[0969] "Please explain the specific processing steps of a system that monitors the congestion status of each vehicle in real time during the morning rush hour, detects suspicious individuals, and reflects user emotions in announcements."

[0970] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0971] Step 1:

[0972] The terminal (train camera) captures real-time images from inside each car every second. The input is raw video data from inside each car, and the output is compressed image data. Specifically, the camera captures images from inside each car at high resolution and compresses the data in JPEG format or similar. The compressed image data is sent to the server.

[0973] Step 2:

[0974] The server receives image data sent from the device. The input is compressed image data, and the output is data ready for analysis. The server inputs the received image data into a trained machine learning model (e.g., a convolutional neural network (CNN)) to analyze the congestion level of each vehicle. Specifically, the server uses the CNN to detect people in the image and calculate the number of people.

[0975] Step 3:

[0976] The server quantifies the degree of congestion in each car based on the image analysis results. The input is the analyzed person detection data, and the output is numerical data indicating the degree of congestion in each car. The server generates specific congestion status information, such as "Car 1 is very crowded" or "Car 6 is relatively empty." The generated information is saved in a database in real time.

[0977] Step 4:

[0978] The server creates text data for announcements based on the generated congestion information. The input is congestion information, and the output is text data for announcements. This text data is then processed by a voice synthesis module and sent to the station announcement system. Specifically, announcements such as "Car 1 is very crowded" and "Car 6 is relatively empty" are made in real time.

[0979] Step 5:

[0980] The terminal (ticket gate system) takes facial images of all people passing through the ticket gate and sends them to the server in real time. The input is the facial image of the person passing through, and the output is the transmitted facial image data. Specifically, the ticket gate system uses a high-resolution camera to take facial images, compresses the data, and sends it to the server.

[0981] Step 6:

[0982] The server analyzes the received facial image data using a facial recognition algorithm and compares it with a database of criminal records. The input is the transmitted facial image data, and the output is the comparison result. If it matches a suspicious person, the server generates warning data and sends it to the ticket gate system and station staff terminals. Specifically, the server uses the dlib library to perform facial recognition and compares it with a database of criminal records.

[0983] Step 7:

[0984] The device (emotion engine) analyzes commuters' facial expressions and recognizes their emotions. The input is facial expression data from each vehicle, and the output is analyzed emotion data. The emotion engine uses open-source facial expression recognition libraries (e.g., OpenCV and Dlib). Specifically, the device acquires facial expression data in real time and identifies emotions such as joy, surprise, sadness, and anger.

[0985] Step 8:

[0986] The server stores and analyzes the recognized emotion data. The input is the analyzed emotion data, and the output is data for adjusting the content of announcements. For example, if many users in a certain car show expressions of anger or dissatisfaction, the server adjusts the announcement content to something more reassuring, such as "Car 6 is relatively empty. Please use it." The text data for the announcement is processed by a speech synthesis module and immediately sent to the station announcement system.

[0987] Specific examples

[0988] For example, if car 1 is extremely crowded during the morning rush hour, image data from the device (in-train camera) is sent to the server, which determines that "car 1 is extremely crowded" and "car 6 is relatively empty." Furthermore, the device (emotion engine) analyzes the user's facial expressions and detects that many users are showing signs of dissatisfaction or stress. The server then adjusts the announcement to "car 6 is relatively empty. Please use it." Based on this announcement, users can head to car 6 and enjoy a comfortable commute. Furthermore, if a suspicious person is detected while passing through the ticket gate, the server immediately issues an alert and notifies the terminal (ticket gate system) and the station staff's terminal. In this way, the system grasps the situation from multiple angles in real time and responds promptly.

[0989] (Application example 2)

[0990] 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."

[0991] Conventional commuter train systems lack the means to grasp congestion status in real time and provide appropriate announcements to passengers. Furthermore, there is a need for security measures such as adjusting announcements to reduce passenger stress and anxiety, and detecting suspicious individuals. This will improve passenger comfort and safety.

[0992] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes image acquisition means for collecting image data of each vehicle, image analysis means for analyzing the image data to determine the congestion level of each vehicle, information generation means for generating information indicating the congestion level of each vehicle based on the congestion level, announcement means for announcing the congestion level based on the information, emotion recognition means for recognizing emotions by analyzing facial expressions of passengers, and announcement adjustment means for adjusting the content of announcements based on the emotions recognized by the emotion recognition means. This makes it possible to grasp the congestion level in real time, detect suspicious individuals early, and adjust announcements taking passenger emotions into consideration.

[0993] The "image acquisition means" is a mechanism that collects image data inside the vehicle using sensors such as cameras installed in each vehicle.

[0994] The "image analysis means" is a device or software for analyzing collected image data and determining the degree of congestion in each vehicle.

[0995] The "information generating means" is a mechanism that generates information indicating the congestion state of each vehicle based on the analyzed congestion degree.

[0996] The "announcement means" is a system that announces the congestion situation to passengers based on the generated congestion situation information.

[0997] The "emotion recognition means" is a device or software that analyzes the facial expressions of passengers captured by cameras inside the vehicle and recognizes their emotions.

[0998] The "announcement adjustment means" is a system that adjusts the content and manner of announcements based on the recognized emotions.

[0999] In this invention, in order to improve the comfort and safety of self-driving vehicles, a system is constructed that grasps the congestion situation in each vehicle and the emotions of passengers in real time, and makes appropriate announcements and displays based on that information.

[1000] The server includes the following means:

[1001] 1. Image acquisition method:

[1002] Real-time images are acquired using cameras installed inside the vehicle.

[1003] Cameras are installed at various locations within the vehicle to provide comprehensive coverage of the situation within each vehicle.

[1004] 2. Image analysis methods:

[1005] The acquired image data is compressed and transmitted to the server.

[1006] The server analyzes the image data using machine learning algorithms, such as convolutional neural networks (CNNs), to determine the level of congestion in each vehicle.

[1007] 3. Information generation means:

[1008] Based on the analysis results, information indicating the congestion status of each car is generated, such as "Car 1 is very crowded" or "Car 6 is relatively empty."

[1009] The generated information is stored in a database on the server in real time.

[1010] 4. Announcement Method:

[1011] Text data for announcements is generated based on the information indicating the congestion state.

[1012] The generated text data is converted into speech via a speech synthesis module and transmitted to the announcement system inside the train or station.

[1013] 5. Emotion recognition means:

[1014] The server uses footage from cameras inside the vehicle to analyze passengers' facial expressions in real time and recognize emotions such as joy, anger, sadness, and surprise.

[1015] Emotion recognition uses techniques such as convolutional neural networks (CNNs) for emotion classification.

[1016] 6. Announcement Coordination Methods:

[1017] Adjust the content and manner of announcements based on the recognized emotional data.

[1018] For example, if many passengers look dissatisfied, an announcement will be made saying, "The next car is available, so please use it."

[1019] As a concrete example, imagine a situation where car 1 is very crowded during the morning rush hour, but car 6 is relatively empty. Image data captured by the camera is analyzed by the server, and information such as "car 1 is very crowded" and "car 6 is relatively empty" is generated. This information is provided to passengers as a voice announcement. Furthermore, if the passenger's facial expression indicates dissatisfaction, the server will make a more considerate announcement.

[1020] An example of a prompt is shown below.

[1021] "We will develop a system that analyzes the congestion situation and passenger emotions in real time from camera footage of autonomous vehicles, and adjusts the content of announcements and displays based on that analysis. For example, if the congestion level is high and passengers are expressing dissatisfaction, the system will announce that the next car is available, improving passenger comfort."

[1022] In this way, it is possible to realize a system that enhances the comfort and safety of autonomous vehicles and improves passenger satisfaction.

[1023] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1024] Step 1:

[1025] The terminal (in-car camera) captures real-time images from inside the car. The camera is installed in a specific position inside each car and captures the overall congestion situation and the facial expressions of passengers. The camera captures image data every second, compresses this data, and sends it to the server.

[1026] Input: Real-time video from inside the vehicle

[1027] Output: Compressed image data

[1028] Step 2:

[1029] The server receives the image data sent from the device. Each received image is analyzed using a machine learning algorithm. Specifically, a convolutional neural network (CNN) model is used to determine the degree of vehicle congestion.

[1030] Input: Compressed image data

[1031] Action: Data extraction and image analysis

[1032] Output: Quantified congestion data

[1033] Step 3:

[1034] The server generates information indicating the congestion status of each car based on the analysis results. For example, if the analysis shows that car 1 is highly crowded, it generates information such as "car 1 is very crowded."

[1035] Input: Data quantifying congestion

[1036] Action: Generate congestion information

[1037] Output: Information showing the occupancy status of each vehicle

[1038] Step 4:

[1039] The server then generates text data for announcements based on the congestion status, which is then converted into speech by a speech synthesis module and sent to the announcement system.

[1040] Input: Information indicating the occupancy status of each vehicle

[1041] Operation: Generation of text data for announcements and speech synthesis

[1042] Output: Announcement audio data

[1043] Step 5:

[1044] The terminals (speakers inside the train and in stations) receive the announcement audio data sent from the server and make the actual announcements, allowing passengers to understand the congestion situation in real time.

[1045] Input: Announcement audio data

[1046] Action: Make an announcement

[1047] Output: Announcement voice

[1048] Step 6:

[1049] The terminal (in-vehicle camera) captures facial expression data of passengers. Similarly, image data is captured every second, compressed, and sent to the server.

[1050] Input: Real-time video of passenger facial expressions

[1051] Output: Compressed facial expression data

[1052] Step 7:

[1053] The server receives the facial expression data and uses a machine learning model for emotion recognition (e.g., emotion classification model) to recognize emotions such as joy, anger, sadness, and surprise.

[1054] Input: Compressed facial expression data

[1055] Action: Facial expression analysis and emotion recognition

[1056] Output: Emotion data for each passenger

[1057] Step 8:

[1058] The server then adjusts the announcement content based on the recognized emotion data. For example, if many passengers show expressions of dissatisfaction or stress, the announcement will be changed to one that promotes a sense of security, such as "The next car is available, so please use it."

[1059] Input: Emotion data for each passenger

[1060] Action: Adjust announcement content

[1061] Output: Adjusted text data for announcement

[1062] Step 9:

[1063] The server converts the adjusted text data for announcement into voice again using the voice synthesis module and transmits it to the announcement system.

[1064] Input: Adjusted text data for announcement

[1065] Operation: Generation of announcement voice data and speech synthesis

[1066] Output: Adjusted announcement audio data

[1067] Step 10:

[1068] The terminals (speakers inside the train and at stations) receive the adjusted announcement audio data sent from the server and make announcements based on it.

[1069] Input: Adjusted announcement audio data

[1070] Action: Announcement is performed after adjustment

[1071] Output: Adjusted announcement audio

[1072] In this way, a system is built that can grasp congestion status in real time and make appropriate announcements based on that information. In addition, by adjusting the content of announcements to take passengers' emotions into consideration, a comfortable riding environment is provided.

[1073] 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.

[1074] 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.

[1075] 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.

[1076] [Fourth embodiment]

[1077] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1078] 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.

[1079] 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).

[1080] 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.

[1081] 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.

[1082] 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).

[1083] 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.

[1084] 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.

[1085] 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.

[1086] 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.

[1087] 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.

[1088] 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.

[1089] 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."

[1090] The present invention provides a system for grasping the congestion status of commuter trains in real time and detecting suspicious individuals. The system includes an image acquisition unit, an image analysis unit, an information generation unit, an announcement unit, a feature detection unit, a comparison unit, and a warning unit.

[1091] Program processing and implementation form

[1092] 1. Image acquisition method

[1093] The terminal (in-train camera) periodically captures image data of each car. For example, a camera installed in each car captures an image of passengers every second and collects this image data.

[1094] 2. Image analysis methods

[1095] The server analyzes the image data sent from the device in real time using a trained machine learning model (e.g., convolutional neural network, CNN).The server uses the image data to determine the degree of congestion in each vehicle and quantifies the number and density of passengers in each vehicle.

[1096] 3. Information generation means

[1097] Based on the analyzed congestion level, the server generates information indicating the congestion status of each car (e.g., "Car 1 is 95% crowded," "Car 6 is 50% crowded," etc.). This information is stored in a database in real time.

[1098] 4. Announcement Methods

[1099] The server sends the generated congestion status information to the station announcement system. This allows station staff or automated announcements to provide users with information such as "Car 1 is very crowded. Car 6 is relatively empty." Users (commuters) can then select the car they want to board based on the announcements.

[1100] 5. Feature detection methods

[1101] The terminal (ticket gate system) captures a facial image of a person passing through the ticket gate, and the captured facial image data is sent to a server in real time.

[1102] 6. Verification Method

[1103] The server uses a facial recognition algorithm to extract features from the received facial image and compares it with a database of criminal records, which contains facial feature data of past criminals.

[1104] 7. Warning measures

[1105] If the server finds a match, it determines the person is suspicious and generates a warning. A warning message is displayed on the terminal (ticket gate system), and a warning notification is simultaneously sent to the station staff's terminal. This alert allows station staff to respond immediately and prevent crimes from occurring.

[1106] Example

[1107] For example, during rush hour on a certain weekday morning, car 1 is very crowded, but car 6 is relatively empty. In this case, image data taken by the device is sent to the server, which analyzes it and determines that "car 1 is 95% crowded" and "car 6 is 50% crowded." Based on this, an announcement is made that "car 6 is relatively empty." The user (commuter) can select car 6 based on this information to enjoy a comfortable commute.

[1108] Furthermore, if a suspicious person is detected while passing through a ticket gate, the server immediately sends an alert, which is displayed on the terminal (ticket gate system) and the station staff's terminal.Station staff can then respond quickly and prevent crimes from occurring.

[1109] Thus, the present invention is an effective means for alleviating congestion on commuter trains and improving safety.

[1110] The processing flow will be explained below.

[1111] Specific processing flow of the program

[1112] 1. Understanding congestion and making announcements

[1113] Step 1: Image acquisition

[1114] The terminal (in-train camera) captures real-time images from inside each car every second. This image data is compressed immediately after capture and sent to the server.

[1115] Step 2: Image reception and analysis

[1116] The server receives the image data sent from the terminal, immediately stores it in a database, and inputs it into the analysis module.

[1117] The server's image analysis module uses machine learning models (such as CNN) to analyze the congestion level of each vehicle, calculating the number of passengers, their density, the number of people standing, and other factors to quantify the level of congestion.

[1118] Step 3: Generate congestion information

[1119] Based on the analysis results, the server quantifies the degree of congestion in each vehicle and classifies it into categories such as "very crowded," "moderately crowded," or "empty" according to certain criteria.

[1120] Step 4: Generate announcement data

[1121] The server generates text data for announcements based on the generated congestion information. For example, "Car 1 is very crowded. Car 6 is relatively empty."

[1122] Step 5: Making the announcement

[1123] The server sends the generated text data to the station's announcement system, where it is processed by a speech synthesis module and an announcement is made.

[1124] Users (commuters) listen to the announcement and select the car they want to get in. For example, if they know that car 1 is crowded, they can choose car 6.

[1125] 2. Criminal Detection and Warning

[1126] Step 1: Image acquisition

[1127] The terminal (ticket gate system) takes facial images of all people passing through the ticket gate, and the captured image data is sent to the server in real time.

[1128] Step 2: Receiving and preprocessing face images

[1129] The server receives the facial image data sent from the ticket gate system, and immediately extracts facial feature points using a feature point extraction algorithm.

[1130] Step 3: Matching facial feature points

[1131] The server compares the minutiae data with a criminal record database. A matching algorithm calculates the degree of match between the minutiae and determines a match if the degree of match exceeds a certain threshold.

[1132] Step 4: Match detection and warning generation

[1133] If a match is found, the server generates alert data, which includes the suspect's characteristics and location.

[1134] Step 5: Sending an alert

[1135] The server sends the generated warning data to the terminal (ticket gate system) and the station staff's terminal. A warning message is displayed on the ticket gate system, and an alert is sent to the station staff's terminal.

[1136] Upon receiving the alert, station staff immediately take steps to deal with suspicious persons.

[1137] The above is the specific operation of each step. This system makes full use of image analysis and facial recognition technology to reduce congestion on commuter trains and improve safety.

[1138] Example 1

[1139] 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."

[1140] On current commuter trains, it is difficult to grasp the congestion status of each car in real time, which results in passengers often boarding cars that are more crowded. Also, if a suspicious person passes through the ticket gate, there are limited ways to quickly deal with the problem, which poses safety issues.

[1141] 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.

[1142] In this invention, the server includes image acquisition means for collecting image data of each vehicle, image analysis means for analyzing the image data to determine the degree of congestion of each vehicle, information generation means for generating information indicating the congestion status of each vehicle based on the congestion status, announcement means for announcing the congestion status based on the information, image acquisition means for taking facial images of people passing through the ticket gate, comparison means for analyzing the facial images to extract feature points and comparing them with a criminal record database, and warning means for generating and notifying warning data if a match is confirmed by the comparison means. This allows passengers to choose vehicles that are less crowded, improving safety.

[1143] The "image acquisition means" is a device for capturing and collecting image data of passengers and passersby inside trains and at ticket gates.

[1144] "Image analysis means" is a technology for analyzing acquired image data and extracting specific information (for example, congestion level or characteristic points).

[1145] An "information generation means" is a process or system for generating useful information (e.g., congestion status) from analyzed data.

[1146] "Announcement means" refers to audio and display devices used to inform passengers and other relevant parties of the situation based on the generated information.

[1147] The "matching means" is a system for comparing the feature points of the acquired facial image with an existing database to confirm a match.

[1148] The "warning means" is a mechanism for generating warning data and notifying relevant parties when a match is confirmed.

[1149] "Crowding level of each car" is an index that indicates the number and density of passengers in each car of a train.

[1150] A "criminal record database" is a database that registers facial images and feature points of people who have a criminal record.

[1151] "Feature detection means" is a technology for analyzing and extracting feature points from the facial image of a person passing through a ticket gate.

[1152] This invention is a system for understanding the congestion situation of commuter trains in real time and detecting suspicious individuals. This system operates through a series of processes: image data acquisition, analysis, and warning notification. Each function is explained in detail below.

[1153] Image Acquisition Method

[1154] The devices (in-train cameras and ticket gate system cameras) periodically capture image data of passengers and passersby and send it to a server. For example, high-resolution cameras (generic name) installed in each car and ticket gate system cameras (generic name) are used. These cameras automatically use a timer function to periodically capture images, store the data in temporary storage, and then send it to the server.

[1155] Image analysis methods

[1156] The server analyzes the received image data in real time. This analysis uses a pre-trained machine learning algorithm (e.g., convolutional neural network, generic name). Software used for the analysis includes TensorFlow (generic name). The received image data is processed by a server equipped with a high-performance GPU (generic name), which quantifies the number of passengers and their density to determine the level of congestion.

[1157] Information generation means

[1158] The server generates information showing the congestion status of each car based on the analyzed congestion level information. For example, information such as "Car 1 is 95% crowded" or "Car 6 is 50% crowded" is generated. This information is saved in a database in real time and kept up to date. MySQL (generic name) or similar is used as the database.

[1159] Announcement Method

[1160] The server sends the generated congestion status information to the announcement system within the station. This allows station staff or automated voice systems to provide users with information such as, "Car 1 is very crowded. Car 6 is relatively empty." Users (passengers) can choose their own car based on this information, allowing them to avoid crowds and enjoy a comfortable commute.

[1161] Feature detection and matching methods

[1162] The terminal (ticket gate system) takes a picture of the face of a person passing through the ticket gate and sends it to the server. The server analyzes the received face image using a facial recognition algorithm (e.g., OpenCV, generic name) and extracts feature points. The extracted feature points are then compared with a database of criminal records.

[1163] warning means

[1164] If the server finds a match, it determines the person is suspicious and generates a warning. The generated warning data is sent in real time to the ticket gate system and station staff terminals, where a warning message is displayed. This allows station staff to respond quickly and prevent crimes from occurring.

[1165] Specific examples

[1166] For example, during rush hour on a certain weekday morning, car 1 is very crowded, but car 6 is relatively empty. In this case, image data taken by the device is sent to the server, which analyzes it and determines that "car 1 is 95% crowded" and "car 6 is 50% crowded." Based on this, an announcement is made that "car 6 is relatively empty." The user (commuter) can select car 6 based on this information to enjoy a comfortable commute.

[1167] Furthermore, if a suspicious person is detected while passing through a ticket gate, the server immediately sends an alert, which is displayed on the ticket gate system and on the station staff's terminals, allowing station staff to respond quickly and prevent crimes from occurring.

[1168] Prompt Sentence Examples

[1169] "Please explain the system that monitors the congestion status of train cars in real time during the weekday morning rush hour and selects the appropriate car. Please also explain the process for detecting suspicious individuals at ticket gates and issuing warnings."

[1170] Thus, the present invention is a very effective means for alleviating congestion on commuter trains and improving safety.

[1171] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1172] Step 1: Image acquisition

[1173] The terminals (in-train cameras and ticket gate system cameras) capture image data of each car and people passing through the ticket gates. The input data are still images taken by the cameras, which are saved in temporary storage. Specifically, the cameras record images every second and temporarily save these images in memory. The saved image data is then sent to the server.

[1174] Step 2: Receiving image data

[1175] The server receives image data sent from the terminal. The input data is an image file taken by the terminal camera. The server receives this image data in real time and temporarily stores it in storage. This storage is used as the input data source for the next analysis process.

[1176] Step 3: Image analysis

[1177] The server analyzes the received image data. The input is the received image files, and the output is the congestion level of each car and facial feature data of people passing through the ticket gates. This analysis uses a trained machine learning algorithm (e.g., convolutional neural network, CNN). The server uses a GPU to process the images at high speed and extracts the number and density of passengers and facial feature points.

[1178] Step 4: Congestion assessment

[1179] The server quantifies the congestion level of each car based on the results of image analysis. The input is the number of passengers and density data obtained from image analysis, and the output is the congestion level of each car (e.g., "Car 1 is 95% crowded," "Car 6 is 50% crowded"). Based on this numerical data, the congestion level of each car is evaluated and information is generated.

[1180] Step 5: Information Generation

[1181] The server creates announcement data based on the generated congestion status information. The input data is the congestion assessment result, and the output is information in the form of a message to be sent to the announcement system. For example, information such as "Car 1 is very crowded. Car 6 is relatively empty" is generated.

[1182] Step 6: Announcement

[1183] The server sends the created announcement data to the announcement system in the station. The input data is the message information for the announcement, and the output is the information to be displayed on the station's speakers or displays. This allows station staff and automated voice systems to provide commuters with congestion information. Users (commuters) can select their train car based on this announcement information.

[1184] Step 7: Extract facial feature points

[1185] The server analyzes the facial image data of people passing through the ticket gate and extracts facial feature points. The input data is the facial image sent from the ticket gate system camera, and the output data is the extracted facial feature points. Here, the facial feature points are extracted using a machine learning algorithm (e.g., OpenCV).

[1186] Step 8: Face data matching

[1187] The server compares the extracted facial feature points with a database of criminal records. The input data is the facial feature points and known face data in the database, and the output data is the comparison result. If a match is confirmed as a result of the comparison, the person is judged to be suspicious.

[1188] Step 9: Generate warning data

[1189] The server generates warning data based on the matching results. The input data is the facial data matching result, and the output data is a warning message. If a suspicious person is detected, the server immediately generates warning data and sends the message to the terminal (ticket gate system or station staff terminal).

[1190] Step 10: Warning Notification

[1191] The terminals (ticket gate system and station staff terminal) receive the warning data sent from the server and display the warning message. The input data is the warning message from the server, and the output is the warning message displayed on the terminal display. Station staff can check this warning and take action immediately.

[1192] The above is a specific flow of each processing step in the system of the present invention, which makes it possible to grasp the congestion situation on commuter trains and detect suspicious individuals.

[1193] (Application example 1)

[1194] 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."

[1195] In modern manufacturing factories, robots are required to efficiently transport raw materials and products, but this is often hindered by congestion within the factory and the presence of obstacles. From a safety perspective, detecting suspicious objects and obstacles is also important, but existing systems are not able to adequately address this. There is a need for a new system that can solve these problems and achieve efficient and safe transportation.

[1196] 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.

[1197] In this invention, the server includes an image acquisition unit for collecting video data of each area, an image analysis unit for analyzing the video data to determine the congestion level of each area, an information generation unit for generating information indicating the congestion level of each area based on the congestion level, a feature detection unit for detecting suspicious objects and obstacles, and a warning unit for issuing a warning about suspicious objects and obstacles detected by the feature detection unit. This makes it possible to grasp the congestion level within the factory in real time and optimize the robot's transportation route. Furthermore, detecting suspicious objects and obstacles can improve safety.

[1198] "Image acquisition means" refers to a device such as a camera mounted on a robot or peripheral equipment, and is a means for collecting image data of an area.

[1199] The "image analysis means" is a means for analyzing collected image data to determine the degree of congestion in an area and the presence or absence of suspicious objects and obstacles. Specifically, it uses machine learning algorithms to analyze images.

[1200] The "information generating means" is a means for generating information indicating the situation of an area based on the degree of congestion obtained by the image analyzing means and the detected objects.

[1201] The "announcement means" is a means for communicating the generated congestion information to robots and operators in the factory by voice or display.

[1202] The "feature detection means" is a means for detecting suspicious objects and obstacles from the collected image data.

[1203] The "warning means" is a means for issuing a warning to notify the operator when a suspicious object or obstacle is detected.

[1204] A system for realizing the present invention is configured as follows.

[1205] First, cameras mounted on the robots and peripheral equipment collect video data of each area in real time. This is the image acquisition method. The collected video data is then sent to a server via a network.

[1206] The server receives the transmitted video data and analyzes it. For the analysis, it uses machine learning algorithms such as convolutional neural networks (CNN). This allows it to determine the degree of congestion in each area and the presence of obstacles and suspicious objects. This is the image analysis method.

[1207] Next, the server generates information indicating the congestion status and safety of each area based on the information obtained by the image analysis means. This is the information generation means.

[1208] The generated information is announced to the user (factory operator or robot). Instructions for the transport route are sent to the robot in real time via the announcement means. Specifically, instructions include "Please select a different route due to high congestion" or "Please proceed along this route."

[1209] Furthermore, the server uses a feature detection means to detect suspicious objects and obstacles. If any are detected, a warning means is activated to issue a warning to the operator. This allows for a quick response and improves factory safety.

[1210] Consider the following scenario as a specific example. For example, if a specific area in a factory is congested, images captured by a camera are sent to a server in real time. The server analyzes the images and determines that the congestion level in that area is 90% or higher, and sends an instruction to the robot to "select an alternative route." If a suspicious object is detected, an alert is immediately issued and a notification is sent to the operator.

[1211] In this way, this invention makes it possible to grasp the congestion situation in a factory in real time and optimize the robot's transportation route. It also improves safety by detecting suspicious objects and obstacles.

[1212] Examples of prompts include:

[1213] "Please tell me how to implement software that analyzes the congestion status and presence of obstacles in a factory in real time. The system uses cameras, robots, and servers, and uses Python, OpenCV, and Keras. Please explain with specific code examples."

[1214] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1215] Step 1:

[1216] The terminal (camera) collects video data from each area in the factory in real time. Specifically, the camera periodically captures images of the situation within the area and acquires the image data. The input is the video data captured by the camera, and the output is the data saved on the terminal.

[1217] Step 2:

[1218] The terminal (camera) transmits the collected video data to the server via the network. The server receives the data and prepares it for the next analysis step. The input is the video data, and the output is the data sent to the server.

[1219] Step 3:

[1220] The server preprocesses the received video data, specifically resizing and normalizing the image. The input is the received video data, and the output is the preprocessed image data.

[1221] Step 4:

[1222] The server inputs the preprocessed image data into a machine learning algorithm to analyze the level of congestion, obstacles, and the presence of suspicious objects. Here, a generative AI model such as a convolutional neural network (CNN) is used. The input is the preprocessed image data, and the output is the analysis results.

[1223] Step 5:

[1224] The server generates information based on the analysis results. Specifically, it generates information indicating the congestion status of each area and the presence of suspicious objects, and stores this information in a database. The input is the analysis results, and the output is the generated information data.

[1225] Step 6:

[1226] The server notifies the users (robots and operators) of the generated information through the announcement means. The robot is informed of the transport route and the operator of safety information. The input is the generated information data and the output is notification data.

[1227] Step 7:

[1228] The device (camera) uses the feature detection means to re-detect suspicious objects and obstacles, and if any abnormalities are found, it sends the data to the server again. The input is the newly captured video data, and the output is the re-analyzed data.

[1229] Step 8:

[1230] The server analyzes the received data again, and if an abnormality is found, it issues a warning to the operator using the warning means. The input is the reanalyzed data, and the output is a warning notification.

[1231] 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.

[1232] The present invention is a system that grasps the congestion status of commuter trains in real time, detects suspicious individuals, and recognizes user emotions to adjust announcements and displays. This system includes image acquisition means, image analysis means, information generation means, announcement means, feature detection means, comparison means, warning means, and an emotion engine.

[1233] Program processing and implementation form

[1234] 1. Understanding congestion and making announcements

[1235] The terminal (in-train camera) captures real-time images from inside each car every second. This image data is compressed immediately after capture and sent to the server.

[1236] The server receives the image data sent from the terminal and analyzes it using a trained machine learning model (such as CNN) to determine the degree of congestion in each vehicle. The analysis results are used to quantify the number and density of passengers in each vehicle.

[1237] The server generates information indicating the congestion status of each car based on the analyzed congestion level. For example, "Car 1 is very crowded" or "Car 6 is relatively empty." This information is saved in a database in real time.

[1238] The server generates text data for announcements based on the congestion information and sends it to the station's announcement system. This data is processed by a voice synthesis module, and the announcement is made. Users (commuters) can select the car they want to board based on this announcement.

[1239] 2. Criminal Detection and Warning

[1240] The terminal (ticket gate system) takes facial images of all people passing through the ticket gate and transmits them to the server in real time.

[1241] The server uses a facial recognition algorithm to compare the received facial image data with a database of criminal records. If a match is found, the person is deemed suspicious and a warning message is generated. The warning message is sent to the terminal (ticket gate system) and the station staff's terminal, where a warning message or alert is immediately displayed.

[1242] 3. User emotion recognition and announcement / display adjustment

[1243] The device (emotion engine) analyzes the commuter's facial expressions and recognizes the user's emotions. The emotion engine acquires facial expression data in real time and identifies emotions such as joy, surprise, sadness, and anger.

[1244] The server stores and analyzes the recognized emotion data. For example, if many users in a certain car show angry or dissatisfied expressions, the content of the announcement regarding the level of congestion can be adjusted as much as possible. An announcement that gives a sense of security, such as "Car 6 is relatively empty. Please use it," can be generated.

[1245] The server then sends the adjusted text data for the announcement to the station announcement system, which then makes the announcement. If necessary, the server can also provide the information to station staff terminals to prompt them to take appropriate action.

[1246] Example

[1247] For example, during the morning rush hour, car 1 may be very crowded, but car 6 may be relatively empty. In this case, image data from the device (in-train camera) is sent to the server, which determines that "car 1 is very crowded" and "car 6 is relatively empty."

[1248] Furthermore, if the device (emotion engine) analyzes the user's facial expressions and detects that many users are showing signs of dissatisfaction or stress, the server will adjust the announcement. An announcement will be made that takes into consideration emotions, such as "Car 6 is relatively empty. Please use it." Users (commuters) can enjoy a comfortable commute by hearing this announcement and heading to car 6.

[1249] Furthermore, if a suspicious person is detected when passing through a ticket gate, the server will immediately issue an alert and notify the terminal (ticket gate system) and the station staff's terminal, enabling a rapid response.

[1250] In this way, the present invention will alleviate congestion, adjust announcements based on emotion recognition, and detect suspicious individuals early, thereby significantly improving the safety and comfort of commuter trains.

[1251] The processing flow will be explained below.

[1252] Specific processing flow of the program

[1253] 1. Understanding congestion and making announcements

[1254] Step 1: Image acquisition

[1255] The terminal (in-train camera) captures real-time images from inside each car every second. The image data is compressed immediately after capture and sent to the server.

[1256] Step 2: Image reception and analysis

[1257] The server receives the image data sent from the terminal, stores it in a database, and inputs the data into the analysis module.

[1258] The server's image analysis module uses a trained machine learning model (such as CNN) to analyze image data and determine the level of congestion in each vehicle. The analysis calculates the number of passengers, their density, the number of people standing, and other factors to quantify the level of congestion.

[1259] Step 3: Generate congestion information

[1260] Based on the analysis results, the server quantifies the degree of congestion in each vehicle and classifies it into categories such as "very crowded," "moderately crowded," or "empty" according to certain criteria.

[1261] Step 4: Generate announcement data

[1262] The server generates text data for announcements based on the generated congestion information. For example, "Car 1 is very crowded. Car 6 is relatively empty."

[1263] Step 5: Making the announcement

[1264] The server sends the generated text data to the station's announcement system, where it is processed by a speech synthesis module and an announcement is made.

[1265] The user (commuter) listens to the announcement and selects the car they will board. They can choose car 6 to avoid the crowded car 1.

[1266] 2. Criminal Detection and Warning

[1267] Step 1: Image acquisition

[1268] The terminal (ticket gate system) takes facial images of all people passing through the ticket gate, and the captured image data is sent to the server in real time.

[1269] Step 2: Receiving and preprocessing face images

[1270] The server receives the facial image data sent from the ticket gate system and inputs it into a facial recognition algorithm that extracts feature points.

[1271] Step 3: Matching facial feature points

[1272] The server compares the extracted feature data with a database of criminal records. The matching algorithm calculates the degree of match between the feature points and determines a match if the degree of match exceeds a certain threshold.

[1273] Step 4: Match detection and warning generation

[1274] If a match is found, the server determines that the individual is suspicious and generates warning data, including the individual's characteristics and location.

[1275] Step 5: Sending an alert

[1276] The server sends the generated warning data to the terminal (ticket gate system) and the station staff's terminal. A warning message is displayed on the ticket gate system, and an alert is sent to the station staff's terminal.

[1277] Upon receiving the alert, station staff immediately take steps to deal with suspicious persons.

[1278] 3. User emotion recognition and announcement / display adjustment

[1279] Step 1: Acquire facial expression data

[1280] The device (emotion engine) captures the facial expressions of commuters in real time, and the captured data is sent to the emotion engine.

[1281] Step 2: Facial Expression Analysis

[1282] The server receives the facial expression data sent from the emotion engine and analyzes it using an emotion recognition algorithm, identifying emotions such as joy, surprise, sadness, and anger.

[1283] Step 3: Storing and analyzing sentiment data

[1284] The server stores the recognized emotion data and performs statistical analysis. If many users in a particular vehicle express anger or dissatisfaction, the emotion data is taken into account in addition to the congestion level.

[1285] Step 4: Adjust your announcement

[1286] The server adjusts the content of the announcement based on the emotional data, for example, "Car 6 is relatively empty. Please use it." to create a sense of security.

[1287] Step 5: Making the announcement

[1288] The server then sends the adjusted text data for the announcement to the station announcement system, which then makes the announcement. If necessary, the information is also provided to station staff's terminals, encouraging them to take appropriate action.

[1289] Users (commuters) can enjoy a more comfortable commute by listening to this announcement and following the emotionally sensitive instructions.

[1290] Example

[1291] For example, during the morning rush hour, car 1 may be very crowded, while car 6 may be relatively empty. In this case, image data from the device (in-train camera) is sent to the server, which then analyzes the data and concludes that "car 1 is very crowded" and "car 6 is relatively empty."

[1292] In addition, the device (emotion engine) analyzes the user's facial expressions, and if many users show expressions of dissatisfaction or stress, the server adjusts the announcement content and makes a reassuring announcement such as, "Car 6 is relatively empty. Please use it." Users (commuters) can hear this announcement and head to car 6 for a comfortable commute.

[1293] Furthermore, if a suspicious person is detected when passing through a ticket gate, the server will immediately issue an alert and notify the terminal (ticket gate system) and the station staff's terminal, allowing for a prompt response.

[1294] As such, the present invention is a system that can alleviate congestion, adjust announcements based on emotion recognition, and detect suspicious individuals early, thereby significantly improving the safety and comfort of commuter trains.

[1295] Example 2

[1296] 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."

[1297] There is a need for a system that can efficiently grasp the congestion situation on commuter trains, detect suspicious individuals early, recognize users' emotions, and provide a comfortable commuting environment. However, conventional technologies have difficulty satisfying all of these requirements simultaneously, and have not been able to sufficiently improve the safety and comfort of commuters.

[1298] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes: image acquisition means for collecting image data of each car; image analysis means for analyzing the image data to determine the congestion level of each car; information generation means for generating information indicating the congestion level of each car based on the congestion level; announcement means for announcing the congestion level based on the information; means for saving the congestion level information in a database in real time; feature detection means for detecting features of a person passing through a ticket gate by referring to a criminal record database; comparison means for comparing the features of the person detected by the feature detection means with data in the criminal record database; warning means for issuing a warning when a match is confirmed by the comparison means; means for transmitting the warning data to the ticket gate system and the station staff terminal; emotion analysis means for analyzing the user's facial expression and recognizing their emotion; and means for adjusting the content of the announcement based on the emotion data recognized by the emotion analysis means. This makes it possible to grasp the congestion level of commuter trains in real time and provide a comfortable commuting environment by detecting suspicious individuals and recognizing the user's emotion.

[1299] "Image acquisition means" refers to a device or function for collecting image data inside a vehicle.

[1300] The "image analysis means" refers to a device or function for analyzing collected image data and determining the degree of congestion inside the vehicle.

[1301] The "information generating means" refers to a device or function for generating information indicating the congestion status of each vehicle based on the analysis results.

[1302] The "announcement means" refers to a device or function for making announcements based on the generated congestion information.

[1303] The "means for storing in a database" refers to a device or function for storing the generated congestion status information in a database in real time.

[1304] "Characteristics detection means" refers to a device or function for detecting the characteristics of a person passing through a ticket gate.

[1305] The "matching means" is a device or function for matching the characteristics of a detected person with data in a criminal record database.

[1306] The "warning means" is a device or function for issuing a warning when a match is confirmed based on the comparison result.

[1307] The "means for transmitting to the ticket gate system and the station staff terminal" refers to a device or function for transmitting warning data to the ticket gate system and the station staff terminal.

[1308] The "emotion analysis means" is a device or function for analyzing the user's facial expression and recognizing the emotion.

[1309] The "means for adjusting the content of the announcement" refers to a device or function for adjusting the content of the announcement based on the recognized emotion data.

[1310] MODE FOR CARRYING OUT THE INVENTION

[1311] The present invention is a system that grasps the congestion status of commuter trains in real time, detects suspicious individuals, and recognizes user emotions to adjust announcements and displays. This system includes image acquisition means, image analysis means, information generation means, announcement means, feature detection means, comparison means, warning means, and an emotion engine.

[1312] 1. Understanding congestion and making announcements

[1313] The device (in-train camera) captures real-time video of each car every second. This image data is compressed and sent to the server. The server then analyzes this image data using a trained machine learning model (e.g., a convolutional neural network (CNN)) to determine the level of congestion in each car. The analysis results are then quantified as the number and density of passengers in each car. The server then uses this information to generate information indicating the congestion level of each car. For example, "Car 1 is very crowded" or "Car 6 is relatively empty." This information is then stored in a database in real time. The server then generates text data for announcements based on the congestion level information and sends it to the station's announcement system. This text data is processed by a speech synthesis module and announced in real time. Users (commuters) can use this information to choose a more comfortable car.

[1314] 2. Suspicious person detection and warning

[1315] The terminal (ticket gate system) takes a facial image of every person passing through the ticket gate and sends it to a server in real time. The server then uses a facial recognition algorithm (for example, dlib, an open-source facial recognition library) to compare the received facial image data with a database of criminal records. If the comparison results in a match with a suspicious individual, the server generates warning data. This warning data is sent to the terminal (ticket gate system) and the station staff's terminal, where a warning message or alert is immediately displayed.

[1316] 3. User emotion recognition and announcement / display adjustment

[1317] The terminal (emotion engine) analyzes commuters' facial expressions and recognizes their emotions. The emotion engine uses open-source facial expression recognition libraries (e.g., OpenCV and Dlib) to acquire facial expression data in real time and identify emotions such as joy, surprise, sadness, and anger. The server stores and analyzes the recognized emotion data. For example, if many users in a certain car show angry or dissatisfied expressions, the announcement content can be adjusted. Specifically, it can be changed to a more reassuring message such as, "Car 6 is relatively empty. Please use it." This adjusted text data for the announcement is also processed by the speech synthesis module and sent to the station announcement system. Furthermore, the server can provide information to station staff terminals as needed to prompt appropriate action.

[1318] Specific examples

[1319] For example, if car 1 is very crowded during the morning rush hour, image data captured by the terminal (in-train camera) is sent to the server. The server analyzes this data and determines that "car 1 is very crowded" or "car 6 is relatively empty." Additionally, if the terminal (emotion engine) analyzes the user's facial expressions and detects that many users are expressing dissatisfaction or stress, the server adjusts the content of the announcement. A reassuring announcement such as "car 6 is relatively empty. Please use it" is made. The user (commuter) can then refer to this announcement and head to car 6, enjoying a comfortable commute. Furthermore, if a suspicious person is detected at the ticket gate, the server immediately generates warning data and sends it to the terminal (ticket gate system) and the station staff's terminal. Thus, the present invention is a system that provides a safe and comfortable commuting environment through commuter train congestion status, emotion recognition, and suspicious person detection.

[1320] Prompt Sentence Examples

[1321] "Please explain the specific processing steps of a system that monitors the congestion status of each vehicle in real time during the morning rush hour, detects suspicious individuals, and reflects user emotions in announcements."

[1322] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1323] Step 1:

[1324] The terminal (train camera) captures real-time images from inside each car every second. The input is raw video data from inside each car, and the output is compressed image data. Specifically, the camera captures images from inside each car at high resolution and compresses the data in JPEG format or similar. The compressed image data is sent to the server.

[1325] Step 2:

[1326] The server receives image data sent from the device. The input is compressed image data, and the output is data ready for analysis. The server inputs the received image data into a trained machine learning model (e.g., a convolutional neural network (CNN)) to analyze the congestion level of each vehicle. Specifically, the server uses the CNN to detect people in the image and calculate the number of people.

[1327] Step 3:

[1328] The server quantifies the degree of congestion in each car based on the image analysis results. The input is the analyzed person detection data, and the output is numerical data indicating the degree of congestion in each car. The server generates specific congestion status information, such as "Car 1 is very crowded" or "Car 6 is relatively empty." The generated information is saved in a database in real time.

[1329] Step 4:

[1330] The server creates text data for announcements based on the generated congestion information. The input is congestion information, and the output is text data for announcements. This text data is then processed by a voice synthesis module and sent to the station announcement system. Specifically, announcements such as "Car 1 is very crowded" and "Car 6 is relatively empty" are made in real time.

[1331] Step 5:

[1332] The terminal (ticket gate system) takes facial images of all people passing through the ticket gate and sends them to the server in real time. The input is the facial image of the person passing through, and the output is the transmitted facial image data. Specifically, the ticket gate system uses a high-resolution camera to take facial images, compresses the data, and sends it to the server.

[1333] Step 6:

[1334] The server analyzes the received facial image data using a facial recognition algorithm and compares it with a database of criminal records. The input is the transmitted facial image data, and the output is the comparison result. If it matches a suspicious person, the server generates warning data and sends it to the ticket gate system and station staff terminals. Specifically, the server uses the dlib library to perform facial recognition and compares it with a database of criminal records.

[1335] Step 7:

[1336] The device (emotion engine) analyzes commuters' facial expressions and recognizes their emotions. The input is facial expression data from each vehicle, and the output is analyzed emotion data. The emotion engine uses open-source facial expression recognition libraries (e.g., OpenCV and Dlib). Specifically, the device acquires facial expression data in real time and identifies emotions such as joy, surprise, sadness, and anger.

[1337] Step 8:

[1338] The server stores and analyzes the recognized emotion data. The input is the analyzed emotion data, and the output is data for adjusting the content of announcements. For example, if many users in a certain car show expressions of anger or dissatisfaction, the server adjusts the announcement content to something more reassuring, such as "Car 6 is relatively empty. Please use it." The text data for the announcement is processed by a speech synthesis module and immediately sent to the station announcement system.

[1339] Specific examples

[1340] For example, if car 1 is extremely crowded during the morning rush hour, image data from the device (in-train camera) is sent to the server, which determines that "car 1 is extremely crowded" and "car 6 is relatively empty." Furthermore, the device (emotion engine) analyzes the user's facial expressions and detects that many users are showing signs of dissatisfaction or stress. The server then adjusts the announcement to "car 6 is relatively empty. Please use it." Based on this announcement, users can head to car 6 and enjoy a comfortable commute. Furthermore, if a suspicious person is detected while passing through the ticket gate, the server immediately issues an alert and notifies the terminal (ticket gate system) and the station staff's terminal. In this way, the system grasps the situation from multiple angles in real time and responds promptly.

[1341] (Application example 2)

[1342] 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."

[1343] Conventional commuter train systems lack the means to grasp congestion status in real time and provide appropriate announcements to passengers. Furthermore, there is a need for security measures such as adjusting announcements to reduce passenger stress and anxiety, and detecting suspicious individuals. This will improve passenger comfort and safety.

[1344] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes image acquisition means for collecting image data of each vehicle, image analysis means for analyzing the image data to determine the congestion level of each vehicle, information generation means for generating information indicating the congestion level of each vehicle based on the congestion level, announcement means for announcing the congestion level based on the information, emotion recognition means for recognizing emotions by analyzing facial expressions of passengers, and announcement adjustment means for adjusting the content of announcements based on the emotions recognized by the emotion recognition means. This makes it possible to grasp the congestion level in real time, detect suspicious individuals early, and adjust announcements taking passenger emotions into consideration.

[1345] The "image acquisition means" is a mechanism that collects image data inside the vehicle using sensors such as cameras installed in each vehicle.

[1346] The "image analysis means" is a device or software for analyzing collected image data and determining the degree of congestion in each vehicle.

[1347] The "information generating means" is a mechanism that generates information indicating the congestion state of each vehicle based on the analyzed congestion degree.

[1348] The "announcement means" is a system that announces the congestion situation to passengers based on the generated congestion situation information.

[1349] The "emotion recognition means" is a device or software that analyzes the facial expressions of passengers captured by cameras inside the vehicle and recognizes their emotions.

[1350] The "announcement adjustment means" is a system that adjusts the content and manner of announcements based on the recognized emotions.

[1351] In this invention, in order to improve the comfort and safety of self-driving vehicles, a system is constructed that grasps the congestion situation in each vehicle and the emotions of passengers in real time, and makes appropriate announcements and displays based on that information.

[1352] The server includes the following means:

[1353] 1. Image acquisition method:

[1354] Real-time images are acquired using cameras installed inside the vehicle.

[1355] Cameras are installed at various locations within the vehicle to provide comprehensive coverage of the situation within each vehicle.

[1356] 2. Image analysis methods:

[1357] The acquired image data is compressed and transmitted to the server.

[1358] The server analyzes the image data using machine learning algorithms, such as convolutional neural networks (CNNs), to determine the level of congestion in each vehicle.

[1359] 3. Information generation means:

[1360] Based on the analysis results, information indicating the congestion status of each car is generated, such as "Car 1 is very crowded" or "Car 6 is relatively empty."

[1361] The generated information is stored in a database on the server in real time.

[1362] 4. Announcement Method:

[1363] Text data for announcements is generated based on the information indicating the congestion state.

[1364] The generated text data is converted into speech via a speech synthesis module and transmitted to the announcement system inside the train or station.

[1365] 5. Emotion recognition means:

[1366] The server uses footage from cameras inside the vehicle to analyze passengers' facial expressions in real time and recognize emotions such as joy, anger, sadness, and surprise.

[1367] Emotion recognition uses techniques such as convolutional neural networks (CNNs) for emotion classification.

[1368] 6. Announcement Coordination Methods:

[1369] Adjust the content and manner of announcements based on the recognized emotional data.

[1370] For example, if many passengers look dissatisfied, an announcement will be made saying, "The next car is available, so please use it."

[1371] As a concrete example, imagine a situation where car 1 is very crowded during the morning rush hour, but car 6 is relatively empty. Image data captured by the camera is analyzed by the server, and information such as "car 1 is very crowded" and "car 6 is relatively empty" is generated. This information is provided to passengers as a voice announcement. Furthermore, if the passenger's facial expression indicates dissatisfaction, the server will make a more considerate announcement.

[1372] An example of a prompt is shown below.

[1373] "We will develop a system that analyzes the congestion situation and passenger emotions in real time from camera footage of autonomous vehicles, and adjusts the content of announcements and displays based on that analysis. For example, if the congestion level is high and passengers are expressing dissatisfaction, the system will announce that the next car is available, improving passenger comfort."

[1374] In this way, it is possible to realize a system that enhances the comfort and safety of autonomous vehicles and improves passenger satisfaction.

[1375] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1376] Step 1:

[1377] The terminal (in-car camera) captures real-time images from inside the car. The camera is installed in a specific position inside each car and captures the overall congestion situation and the facial expressions of passengers. The camera captures image data every second, compresses this data, and sends it to the server.

[1378] Input: Real-time video from inside the vehicle

[1379] Output: Compressed image data

[1380] Step 2:

[1381] The server receives the image data sent from the device. Each received image is analyzed using a machine learning algorithm. Specifically, a convolutional neural network (CNN) model is used to determine the degree of vehicle congestion.

[1382] Input: Compressed image data

[1383] Action: Data extraction and image analysis

[1384] Output: Quantified congestion data

[1385] Step 3:

[1386] The server generates information indicating the congestion status of each car based on the analysis results. For example, if the analysis shows that car 1 is highly crowded, it generates information such as "car 1 is very crowded."

[1387] Input: Data quantifying congestion

[1388] Action: Generate congestion information

[1389] Output: Information showing the occupancy status of each vehicle

[1390] Step 4:

[1391] The server then generates text data for announcements based on the congestion status, which is then converted into speech by a speech synthesis module and sent to the announcement system.

[1392] Input: Information indicating the occupancy status of each vehicle

[1393] Operation: Generation of text data for announcements and speech synthesis

[1394] Output: Announcement audio data

[1395] Step 5:

[1396] The terminals (speakers inside the train and in stations) receive the announcement audio data sent from the server and make the actual announcements, allowing passengers to understand the congestion situation in real time.

[1397] Input: Announcement audio data

[1398] Action: Make an announcement

[1399] Output: Announcement voice

[1400] Step 6:

[1401] The terminal (in-vehicle camera) captures facial expression data of passengers. Similarly, image data is captured every second, compressed, and sent to the server.

[1402] Input: Real-time video of passenger facial expressions

[1403] Output: Compressed facial expression data

[1404] Step 7:

[1405] The server receives the facial expression data and uses a machine learning model for emotion recognition (e.g., emotion classification model) to recognize emotions such as joy, anger, sadness, and surprise.

[1406] Input: Compressed facial expression data

[1407] Action: Facial expression analysis and emotion recognition

[1408] Output: Emotion data for each passenger

[1409] Step 8:

[1410] The server then adjusts the announcement content based on the recognized emotion data. For example, if many passengers show expressions of dissatisfaction or stress, the announcement will be changed to one that promotes a sense of security, such as "The next car is available, so please use it."

[1411] Input: Emotion data for each passenger

[1412] Action: Adjust announcement content

[1413] Output: Adjusted text data for announcement

[1414] Step 9:

[1415] The server converts the adjusted text data for announcement into voice again using the voice synthesis module and transmits it to the announcement system.

[1416] Input: Adjusted text data for announcement

[1417] Operation: Generation of announcement voice data and speech synthesis

[1418] Output: Adjusted announcement audio data

[1419] Step 10:

[1420] The terminals (speakers inside the train and at stations) receive the adjusted announcement audio data sent from the server and make announcements based on it.

[1421] Input: Adjusted announcement audio data

[1422] Action: Announcement is performed after adjustment

[1423] Output: Adjusted announcement audio

[1424] In this way, a system is built that can grasp congestion status in real time and make appropriate announcements based on that information. In addition, by adjusting the content of announcements to take passengers' emotions into consideration, a comfortable riding environment is provided.

[1425] 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.

[1426] 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.

[1427] 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.

[1428] 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.

[1429] 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.

[1430] 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.

[1431] 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).

[1432] 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.

[1433] 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."

[1434] 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.

[1435] 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).

[1436] 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.

[1437] 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.

[1438] 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.

[1439] 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.

[1440] 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.

[1441] 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.

[1442] 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.

[1443] 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.

[1444] 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.

[1445] 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.

[1446] The following is further disclosed regarding the above embodiment.

[1447] (Claim 1)

[1448] image acquisition means for collecting image data of each vehicle;

[1449] image analysis means for analyzing the image data to determine the degree of congestion of each vehicle;

[1450] an information generating means for generating information indicating the congestion state of each vehicle based on the congestion degree;

[1451] an announcing means for announcing the congestion situation based on the information;

[1452] A system including:

[1453] (Claim 2)

[1454] a feature detection means for detecting features of a person passing through the ticket gate by referring to a criminal record database;

[1455] a comparison means for comparing the characteristics of the person detected by the characteristic detection means with data in a criminal record database;

[1456] a warning means for issuing a warning when a match is confirmed by the checking means;

[1457] The system of claim 1 further comprising:

[1458] (Claim 3)

[1459] The image analysis means analyzes the image data using a machine learning algorithm to determine the degree of congestion.

[1460] 10. The system of claim 1.

[1461] "Example 1"

[1462] (Claim 1)

[1463] image acquisition means for collecting image data of each vehicle;

[1464] image analysis means for analyzing the image data to determine the degree of congestion of each vehicle;

[1465] an information generating means for generating information indicating the congestion state of each vehicle based on the congestion degree;

[1466] an announcing means for announcing the congestion situation based on the information;

[1467] image acquisition means for capturing a facial image of a person passing through a ticket gate;

[1468] a matching means for analyzing the facial image, extracting features, and matching the extracted features with a database of criminal records;

[1469] a warning means for generating and notifying warning data when a match is confirmed by the checking means;

[1470] A system including:

[1471] (Claim 2)

[1472] The system according to claim 1, further comprising a characteristic detection means for detecting the characteristics of a person passing through the ticket gate by referring to the criminal record database, and a warning means for issuing a warning based on the match detection result by the comparison means.

[1473] (Claim 3)

[1474] 2. The system according to claim 1, wherein the image analysis means analyzes the image data using a machine learning algorithm to determine the degree of congestion.

[1475] "Application Example 1"

[1476] (Claim 1)

[1477] image acquisition means for collecting video data for each area;

[1478] image analysis means for analyzing the video data to determine the degree of congestion in each area;

[1479] an information generating means for generating information indicating the congestion state of each area based on the congestion degree;

[1480] an announcing means for announcing the congestion situation based on the information;

[1481] a feature detection means for detecting suspicious objects and obstacles;

[1482] a warning means for issuing a warning about a suspicious object or an obstacle detected by the feature detection means;

[1483] A system including:

[1484] (Claim 2)

[1485] a feature detection means for detecting features of a person passing through the ticket gate by referring to a criminal record database;

[1486] a comparison means for comparing the characteristics of the person detected by the characteristic detection means with data in a criminal record database;

[1487] a warning means for issuing a warning when a match is confirmed by the checking means;

[1488] The system of claim 1 further comprising:

[1489] (Claim 3)

[1490] The image analysis means analyzes the video data using a machine learning algorithm to determine the degree of congestion and the presence or absence of suspicious objects.

[1491] 10. The system of claim 1.

[1492] "Example 2: Combining Emotion Engines"

[1493] (Claim 1)

[1494] image acquisition means for collecting image data of each vehicle;

[1495] image analysis means for analyzing the image data to determine the degree of congestion of each vehicle;

[1496] an information generating means for generating information indicating the congestion state of each vehicle based on the congestion degree;

[1497] an announcing means for announcing the congestion situation based on the information;

[1498] means for storing the congestion status information in a database in real time;

[1499] A system including:

[1500] (Claim 2)

[1501] a feature detection means for detecting features of a person passing through the ticket gate by referring to a criminal record database;

[1502] a comparison means for comparing the characteristics of the person detected by the characteristic detection means with data in a criminal record database;

[1503] a warning means for issuing a warning when a match is confirmed by the checking means;

[1504] means for transmitting the warning data to the ticket gate system and a station staff terminal;

[1505] The system of claim 1 further comprising:

[1506] (Claim 3)

[1507] emotion analysis means for analyzing a user's facial expression and recognizing emotions;

[1508] means for adjusting the content of the announcement based on the emotion data recognized by the emotion analysis means;

[1509] The system of claim 1 further comprising:

[1510] "Application example 2 when combining emotion engines"

[1511] (Claim 1)

[1512] image acquisition means for collecting image data of each vehicle;

[1513] image analysis means for analyzing the image data to determine the degree of congestion of each vehicle;

[1514] an information generating means for generating information indicating the congestion state of each vehicle based on the congestion degree;

[1515] an announcing means for announcing the congestion situation based on the information;

[1516] emotion recognition means for analyzing the facial expressions of passengers to recognize their emotions;

[1517] an announcement adjustment means for adjusting the content of the announcement based on the emotion recognized by the emotion recognition means;

[1518] A system including:

[1519] (Claim 2)

[1520] a feature detection means for detecting features of a person passing through the ticket gate by referring to a criminal record database;

[1521] a comparison means for comparing the characteristics of the person detected by the characteristic detection means with data in a criminal record database;

[1522] a warning means for issuing a warning when a match is confirmed by the checking means;

[1523] The system of claim 1 further comprising:

[1524] (Claim 3)

[1525] The image analysis means analyzes the image data using a machine learning algorithm to determine the degree of congestion.

[1526] 10. The system of claim 1. [Explanation of symbols]

[1527] 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. image acquisition means for collecting image data of each vehicle; image analysis means for analyzing the image data to determine the degree of congestion of each vehicle; an information generating means for generating information indicating the congestion state of each vehicle based on the congestion degree; an announcing means for announcing the congestion situation based on the information; A system including:

2. a feature detection means for detecting features of a person passing through the ticket gate by referring to a criminal record database; a comparison means for comparing the characteristics of the person detected by the characteristic detection means with data in a criminal record database; a warning means for issuing a warning when a match is confirmed by the checking means; The system of claim 1 further comprising:

3. The image analysis means analyzes the image data using a machine learning algorithm to determine the degree of congestion. The system of claim 1 .

Images