Inspection support system, inspection support method, and program

Abstract

The present invention provides an inspection support system, an inspection support method, and a program that are capable of accurately estimating whether a person carries or uses a crime-related substance. An inspection support system (1) comprises an information acquisition unit (21) and an estimation unit (23). The information acquisition unit (21) acquires person information pertaining to the state of a person (H1). The estimation unit (23) estimates the possibility that the person (H1) carries or uses a crime-related substance on the basis of a change in the state of the person (H1) in response to an inspection notification indicating that an inspection for detecting the substance is to be performed.

Description

Inspection support system, inspection support method, and program

[0001] The present disclosure relates to an inspection support system, an inspection support method, and a program.

[0002] The abnormal person prediction system of Patent Document 1 predicts whether a person to be monitored within a predetermined area is an abnormal person. The abnormal person prediction system includes a camera means, a behavior pattern storage means, and an abnormal person prediction means. The camera means is installed within the predetermined area and captures images of the person to be monitored. The behavior pattern storage means stores normal behavior patterns or abnormal behavior patterns. The abnormal person prediction means predicts an abnormal person based on the behavior patterns.

[0003] Although the abnormal person prediction system of Patent Document 1 can predict people who are clearly behaving abnormally, it has difficulty detecting people who intentionally behave normally while possessing or using criminal substances such as drugs and explosives.

[0004] Patent No. 7009021

[0005] An object of the present disclosure is to provide an inspection support system, an inspection support method, and a program that can accurately estimate whether a person possesses or uses a substance related to a crime.

[0006] According to one aspect of the present disclosure, there is provided an inspection support system including an information acquisition unit and an estimation unit. The information acquisition unit acquires person information relating to a person's condition. The estimation unit estimates a possibility that the person possesses or uses a substance related to a crime based on a change in the person's condition in response to an inspection notification indicating that an inspection to detect the substance will be conducted.

[0007] According to one aspect of the present disclosure, there is provided an inspection assistance method including: acquiring information about a person's condition in the information acquiring step; and estimating a possibility that the person possesses or uses a substance related to a crime based on a change in the person's condition in response to an inspection notification indicating that an inspection for detecting the substance will be conducted.

[0008] A program according to one aspect of the present disclosure causes a computer system to execute the above-described inspection assistance method.

[0009] FIG. 1 is a block diagram showing a test support system according to a first embodiment. FIG. 2 is an external view showing a portable odor sensor according to a first modified example of the first embodiment. FIG. 3 is a block diagram showing a test support system according to a second modified example of the first embodiment. FIG. 4 is an external view showing a fingerprint authentication device according to a second modified example of the first embodiment. FIG. 5 is a schematic diagram showing a test support system according to a third modified example of the first embodiment. FIG. 6 is a schematic diagram showing a test support system according to a fourth modified example of the first embodiment. FIG. 7 is a flowchart showing a test support method according to the first embodiment. FIG. 8 is a block diagram showing a test support system according to a second embodiment. FIG. 9 is a schematic diagram showing a location where a test notification related to the test support system of the same is made. FIG. 10 is a plan view showing a movement trajectory of a person to be tested in the test support system of the same. FIG. 11 is a flowchart showing an example of processing by the test support system of the same.

[0010] The present embodiment generally relates to an inspection support system, an inspection support method, and a program. More specifically, the present embodiment relates to an inspection support system, an inspection support method, and a program for detecting a person who possesses or uses a substance related to a crime.

[0011] The embodiments and modifications described below are merely examples of the present disclosure, and the present disclosure is not limited to these embodiments and modifications. Various modifications other than these embodiments and modifications are possible depending on the design, etc., as long as they do not deviate from the technical concept of the present disclosure.

[0012] (Embodiment 1) (1) Overview At airports, train stations, stadiums, event venues, lecture halls, and the like, odor sensors are sometimes used to check whether visitors are in possession of or using substances related to crimes, such as drugs or explosives. The odor sensor detects volatile components of drugs, explosives, and the like to determine whether a person is in possession of or using substances related to crimes. However, inspections using odor sensors require the odor sensor to be placed close to the clothing or belongings of the person being inspected, which makes the inspection time-consuming.

[0013] Furthermore, although the abnormal person prediction system of Patent Document 1 mentioned above can predict people who are clearly behaving abnormally, it has difficulty detecting people who intentionally behave normally while possessing or using criminally related substances such as drugs and explosives.

[0014] 1 includes an information acquisition unit 21 and an estimation unit 23. The information acquisition unit 21 acquires person information relating to the state of a person H1. The estimation unit 23 estimates the possibility that the person H1 possesses or uses a substance related to a crime, based on a change in the state of the person H1 in response to the inspection notification. The inspection notification indicates that an inspection to detect a substance related to a crime will be conducted.

[0015] More specifically, the inspection notification indicates that an inspection to detect a substance related to a crime will be conducted on the person H1. The estimation unit 23 estimates the possibility that the person H1 possesses or uses a substance related to a crime, based on a change in the state of the person H1 before and after the person H1 recognizes the inspection notification.

[0016] In the following, a person who possesses or uses a substance involved in a crime will also be referred to as an "offender." A substance involved in a crime will also be referred to as a "violating substance."

[0017] If person H1 (the violator) becomes agitated upon recognizing the test notice, the person information regarding person H1's condition will change before and after recognizing the test notice. That is, if person H1 possesses or uses illegal substances such as drugs or explosives, he or she will become agitated upon recognizing the test notice. Even if person H1 intentionally attempts to behave normally, this agitation will manifest as a change in person information about person H1. Therefore, the test support system 1 having the above-described configuration can accurately estimate whether person H1 possesses or uses illegal substances by estimating the likelihood that person H1 possesses or uses illegal substances based on the change in person information before and after person H1 recognizes the test notice. Furthermore, the use of such a test support system 1 improves the arrest rate for crimes involving possession or use of illegal substances.

[0018] Violative substances include drugs and / or explosives. Violative substances may be, for example, substances related to smuggling or terrorism. Violative substances may also be substances other than drugs and explosives.

[0019] (2) Details The inspection support system 1 shown in Fig. 1 performs inspections for illegal substances by estimating the possibility that a person H1 possesses or uses a illegal substance at facilities such as airports, train stations, stadiums, event venues, and lecture halls. Inspections are performed on person H1 present at inspection points R1, such as airport security checkpoints, immigration or departure checkpoints, station ticket gates or security checkpoints, and entrances to other facilities. Inspections are also performed during police investigations and checkpoints.

[0020] The examination support system 1 includes a calculation unit 2, a human information generation unit 3, a display unit 4, a notification unit 5, and a storage unit 11.

[0021] (2.1) Display Unit The display unit 4 visually displays the text of the inspection notification to the person H1 present at the inspection site R1. That is, the inspection notification is visually displayed to the person H1. The inspection notification indicates that an inspection to detect criminal substances will be conducted on the person H1. In particular, the inspection notification preferably indicates that an inspection to detect criminal substances by analyzing volatile components will be conducted on the person H1. The inspection notification may also indicate that an inspection to detect criminal substances will be conducted on the person H1 by physical examination, X-ray images, or using a sniffer dog, etc.

[0022] Specifically, the display unit 4 includes a display device (such as a liquid crystal display device or an organic EL display device) or a projector device installed at the inspection site R1. The inspection notice is, for example, text such as "We will use an odor sensor to inspect for the presence or absence of illegal substances (substances related to crimes)." In addition to the inspection notice text, the inspection notice may include a symbol representing the inspection or visual effects such as color changes, flashing, or movement within the image to attract the attention of the person H1.

[0023] The display unit 4 is installed in a position that is visible to a person H1 present at the inspection site R1. After entering the inspection site R1, the person H1 sees the inspection notice displayed on the display unit 4 and becomes aware that an inspection will be conducted.

[0024] The display unit 4 may be a display device of a signage system installed at the inspection site R1 or a display device of a kiosk terminal. Alternatively, the display unit 4 may be a poster on which the inspection notification text is printed.

[0025] (2.2) Person Information Generator The person information generator 3 generates person information regarding the state of the person H1 present in the inspection site R1. The state of the person H1 includes at least one of the behavior of the person H1 and the biometric information of the person H1.

[0026] In this embodiment, the human information generation unit 3 includes an imaging device 31, which is a live camera that captures real-time video. The imaging device 31 includes the inspection site R1 in its imaging range, and by capturing an image of the inspection site R1, generates a video capturing the behavior of a person H1 present at the inspection site R1 as human information. That is, the human information generated by the imaging device 31 includes a video capturing the behavior of the person H1 as the state of the person H1. The imaging device 31 outputs the video, which is human information related to the state of the person H1, to the information acquisition unit 21 of the calculation unit 2. The video captures the person H1 entering the inspection site R1, shifting his / her gaze to the display unit 4 and recognizing the inspection notice displayed on the display unit 4, as well as the person's behavior before and after recognizing the inspection notice. The video may be either a color image or a monochrome image.

[0027] (2.3) Calculation Unit The calculation unit 2 includes an information acquisition unit 21, a timing detection unit 22, an estimation unit 23, a result output unit 24, a setting unit 25, and a display control unit 26.

[0028] The calculation unit 2 preferably includes a computer system. The computer system is primarily composed of a processor and memory as hardware. At least a portion of the functions of the calculation unit 2 are realized by the processor executing a program stored in the memory of the computer system. The program may be pre-recorded in the memory of the computer system, provided via a telecommunications line, or provided on a non-transitory recording medium readable by the computer system, such as a memory card, optical disk, or hard disk drive. The processor of the computer system is composed of one or more electronic circuits including a semiconductor integrated circuit (IC) or a large-scale integrated circuit (LSI). The integrated circuits, such as ICs and LSIs, are referred to by different names depending on the degree of integration, and include integrated circuits called system LSIs, very large-scale integrations (VLSIs), or ultra-large-scale integrations (ULSIs). Furthermore, field-programmable gate arrays (FPGAs), which are programmed after the LSI is manufactured, or logic devices that allow the reconfiguration of internal connections or circuit partitions within the LSI, can also be used as processors. The electronic circuits may be integrated into one chip or distributed across multiple chips. The chips may be integrated into one device or distributed across multiple devices. The computer system referred to here includes a microcontroller having one or more processors and one or more memories. Therefore, the microcontroller is also composed of one or more electronic circuits including a semiconductor integrated circuit or a large-scale integrated circuit.

[0029] The computer system is not limited to a single computer device, but may be realized by a plurality of computer devices linked to each other. The computer system may also be constructed as a cloud computing system.

[0030] (2.3.1) Information Acquisition Unit The information acquisition unit 21 acquires person information.

[0031] Specifically, the information acquisition unit 21 acquires, as human information, a video of the inspection site R1 from the imaging device 31, which is the human information generation unit 3. That is, the information acquisition unit 21 acquires a video of the behavior of a person H1 present at the inspection site R1 from the imaging device 31. Then, the information acquisition unit 21 passes the video of the behavior of the person H1 to the timing detection unit 22 and the estimation unit 23.

[0032] Communication between the imaging device 31 and the information acquisition unit 21 is performed by wireless communication or wired communication. The wireless communication is wireless communication that complies with standards such as Wi-Fi (registered trademark), Bluetooth (registered trademark), or low-power wireless communication that does not require a license (specified low-power wireless). The wired communication is wired communication via, for example, a twisted pair cable, a dedicated communication line, or a LAN (Local Area Network) cable.

[0033] (2.3.2) Timing detection unit The timing detection unit 22 detects the recognition timing, which is the timing at which the person H1 recognizes the display of the test notification, based on a video of the person H1. The timing detection unit 22 can easily detect the recognition timing by detecting the recognition timing based on the video.

[0034] Specifically, the timing detection unit 22 receives a video capturing the behavior of person H1 present at the inspection site R1 from the information acquisition unit 21, and detects the recognition timing, which is the timing at which person H1 recognizes the display of the inspection notification, based on the line of sight of person H1 appearing in the video. The timing detection unit 22 passes the detection result of the recognition timing to the estimation unit 23. That is, the timing detection unit 22 can detect the recognition timing within the video. By detecting the recognition timing based on the line of sight, the timing detection unit 22 can accurately detect the recognition timing.

[0035] The timing detection unit 22 applies image recognition technology to the video to analyze the direction of the gaze of the person H1 appearing in the video. The timing detection unit 22 then detects the recognition timing by determining whether the gaze of the person H1 is directed toward the display unit 4 and whether the person H1 recognizes the display of the test notification. The image recognition technology preferably uses an image recognition model constructed by machine learning, specifically a neural network model constructed by deep learning.

[0036] Furthermore, the timing detection unit 22 may analyze the gaze direction of the person H1 appearing in the video by applying a pattern recognition technique to the video. For example, if the gaze of the person H1 appearing in the video matches the display unit 4 and the state in which the gaze matches the display unit 4 continues for a certain period of time or more, the timing detection unit 22 determines that the person H1 has recognized the display of the test notification.

[0037] (2.3.3) Estimation Unit The estimation unit 23 estimates the possibility that the person H1 possesses or uses a violative substance based on the change in the state of the person H1 before and after the person H1 recognizes the display of the test notification.

[0038] Specifically, the estimation unit 23 receives a video of the behavior of the person H1 present at the inspection site R1 from the estimation unit 23, and analyzes the behavior of the person H1 captured in the video as a human state by applying image recognition technology to the video. The behavior includes, for example, the person H1's facial expression, facial direction, gaze, gait, limb movement, finger movement, etc. Then, the estimation unit 23 estimates the possibility that the person H1 possesses or uses a violating substance based on changes in the person H1's behavior before and after the person H1 recognizes the display of the inspection notification.

[0039] In this embodiment, the estimation unit 23 estimates the possibility that the person H1 possesses or uses a violating substance based on fluctuations in the behavior of the person H1 before and after the recognition timing detected by the timing detection unit 22. The estimation unit 23 calculates the possibility that the person H1 possesses or uses a violating substance as an estimated probability. The higher the possibility that the person H1 possesses or uses a violating substance, the higher the estimated probability. The estimated probability is set, for example, to a value in the range of 0 to 1, or a value in the range of 0% to 100%.

[0040] If person H1 possesses or uses a violating substance, he or she will become agitated when he or she recognizes the display of the test notification on the display unit 4. Even if he or she intentionally tries to behave normally, this agitation will manifest as a change in person H1's behavior. For example, when person H1 recognizes the display of the test notification, he or she may change from relaxed behavior to awkward behavior, stiffen his or her facial expression, blink more frequently, look around frequently, suddenly bow his or her head, or try to quickly leave the testing site R1. In other words, when person H1 possesses or uses a violating substance recognizes the display of the test notification, person H1's behavior is likely to change from normal behavior to suspicious behavior. Therefore, the estimation unit 23 estimates the estimation probability based on the change in person H1's behavior before and after the recognition timing.

[0041] The estimation unit 23 analyzes the behavior of the person H1 appearing in the video by applying image recognition technology to the video. The image recognition technology preferably uses an image recognition model constructed by machine learning, specifically a neural network model constructed by deep learning. In the learning process for constructing the image recognition model, supervised learning is performed using training data such as videos of people who begin behaving suspiciously after recognizing the display of an inspection notification on the display unit 4 and videos of people whose behavior remains unchanged before and after recognizing the display of an inspection notification on the display unit 4. The image recognition model constructed by learning then estimates an estimation probability based on the degree of variation in the behavior of the person H1 before and after the recognition timing. The greater the degree of variation in the behavior of the person H1, the higher the estimation probability.

[0042] As described above, the estimation unit 23 can accurately estimate whether person H1 possesses or uses a violating substance by estimating the estimation probability based on changes in person H1's behavior before and after he or she recognizes the display of the test notification.

[0043] The timing detection unit 22 and the estimation unit 23 may be implemented by a single image processing unit. In this case, the image processing unit may use a single image recognition model to perform the process of detecting the recognition timing by analyzing the line of sight of the person H1 and the process of estimating the estimation probability by analyzing the behavior of the person H1.

[0044] (2.3.4) Result Output Unit The result output unit 24 outputs the estimation result of the estimation unit 23 to the notification unit 5.

[0045] Specifically, the result output unit 24 creates test data based on the estimation result of the estimation unit 23 and outputs the test data to the notification unit 5. The test data includes at least one of image data and audio data indicating the estimation result of the estimation unit 23. In the present embodiment, the test data includes both image data and audio data indicating the estimation result of the estimation unit 23.

[0046] The result output unit 24 may include the estimated probability estimated by the estimation unit 23 in the test data, or may include binary data based on the estimated probability in the test data.

[0047] When the test data includes an estimated probability, the test data expresses the estimated probability as a numerical value in the range of 0 to 1, or in the range of 0% to 100%.

[0048] When binary data based on the estimated probability is included in the test data, the result output unit 24 includes binary data indicating "no violating substance" if the estimated probability is less than a predetermined threshold, and "violating substance present" if the estimated probability is greater than or equal to the threshold.

[0049] (2.3.5) Setting Unit The setting unit 25 sets the odor sensor. However, it is not necessary to use the odor sensor to inspect the person H1. In this embodiment, it is assumed that the odor sensor is not used, and therefore the setting unit 25 is not used. The setting unit 25 will be described in the first modified example below.

[0050] (2.3.6) Display Control Unit In this embodiment, the display control unit 26 is not used. The display control unit 26 will be described in the first modified example below.

[0051] (2.4) Notification Unit The notification unit 5 includes at least one of a display device that visually notifies the user of the test data and a speaker device that audibly notifies the user of the test data.

[0052] For example, the notification unit 5 is configured with a personal computer, a dedicated terminal, a smartphone, or a tablet terminal managed by an inspector. The notification unit 5 can communicate with the result output unit 24 of the calculation unit 2 via wireless communication or wired communication. The wireless communication is wireless communication that complies with standards such as Wi-Fi (registered trademark), Bluetooth (registered trademark), or low-power wireless (specified low-power wireless) that does not require a license. The wired communication is wired communication via, for example, a twisted pair cable, a dedicated communication line, or a LAN (Local Area Network) cable.

[0053] When the notification unit 5 receives the test data from the result output unit 24, it displays an image notifying the inspector of the estimated result based on the image data included in the test data. In addition, the notification unit 5 emits a sound notifying the inspector of the estimated result based on the audio data included in the test data.

[0054] Therefore, the inspector can take appropriate action against the person H1 by receiving notification from the notification unit 5 of the estimated probability that the person H1 possesses or uses a violating substance. For example, the inspector can call out to the person H1 whose estimated probability is equal to or greater than a predetermined value and conduct a further inspection. Alternatively, the inspector can call out to the person H1 who has been determined to have a "violating substance" and conduct a further inspection.

[0055] If multiple people H1 appear in the video captured by the imaging device 31, the timing detection unit 22 detects the recognition timing of each of the multiple people H1. Then, the estimation unit 23 estimates the possibility that each of the multiple people H1 possesses or uses a violating substance.

[0056] (2.5) Storage Unit The storage unit 11 is a storage device configured with a hard disk drive (HDD) or a solid state drive (SSD), etc. The storage unit 11 stores the person information acquired by the information acquisition unit 21. For example, the storage unit 11 stores a video as the person information. The video is a video capturing the behavior of person H1. The video stored in the storage unit 11 can be played back, for example, on an information terminal such as a personal computer owned by the inspector. Therefore, the inspector can check the behavior of person H1 by watching the video.

[0057] (3) First Modification Hereinafter, a first modification of the first embodiment will be described. Hereinafter, the configuration of the first embodiment described above will be referred to as a first basic example.

[0058] The inspection support system 1 may be incorporated into a portable odor sensor 8 shown in FIG. 2 . In this case, the imaging device 31 and display unit 4 are arranged on the surface of a case 81 of the odor sensor 8, and the calculation unit 2 is housed inside the case 81. The odor sensor 8 has the function of determining the presence or absence of a violating substance by detecting the volatile components of the violating substance. The odor sensor 8 has a sensitive element whose electrical characteristic value changes depending on the volatile components of the violating substance, and determines the presence or absence of the violating substance based on the electrical characteristic value of the sensitive element. The result of the odor sensor 8's determination of the presence or absence of the violating substance is displayed on the display unit 4.

[0059] First, inspector H2 holds the odor sensor 8 and brings it close to person H1 while capturing an image of person H1 with the imaging device 31 of the odor sensor 8. Then, while the imaging device 31 is capturing an image of person H1, the display unit 4 displays the text of the test notification. By recognizing the test notification displayed on the display unit 4 and further bringing the odor sensor 8 closer, if person H1 possesses or uses a violating substance, changes in behavior will become more apparent. In this modified example, the result output unit 24 of the calculation unit 2 outputs the test data to the display unit 4. The display unit 4 displays an image notifying inspector H2 of the estimated result based on the image data included in the test data. In other words, the display unit 4 also functions as a display device for the notification unit 5 (see FIG. 1) described above.

[0060] The display unit 4 may be used as a display device that displays to the person H1 that detection of a violating substance is being performed. That is, the test notification may include displaying on the display device that detection of a violating substance is being performed for the person H1. This allows the person H1 to recognize that detection of a violating substance is being performed, and if the person H1 is a violator, the person H1 is more likely to become agitated and change their behavior, thereby improving the estimation accuracy of the estimation unit 23. For example, when the odor sensor 8 is brought close to the person H1, the display unit 4 displays text such as "detecting" to indicate that detection of a violating substance is being performed. Alternatively, the display unit 4 displays the measurement value measured by the odor sensor 8 as a numerical value or a graph to indicate that detection of a violating substance is being performed. The numerical value or graph may change in real time depending on the measurement status of the odor sensor 8.

[0061] The testing support system 1 includes a setting unit 25 (see FIG. 1 ). The setting unit 25 changes the settings of the odor sensor 8 based on the estimation result of the estimation unit 23. The odor sensor 8 is a sensor that detects violating substances by analyzing volatile components. The setting unit 25 sets, for example, the sensitivity of the odor sensor 8 to the violating substance, the test time (the length of time for the odor sensor 8 to perform the test), and a judgment threshold. The odor sensor 8 has a sensing element whose electrical characteristic value changes depending on the volatile components of the violating substance, and determines the presence or absence of the violating substance by comparing the electrical characteristic value with the judgment threshold. For example, if the estimation result of the estimation unit 23 is "presence of a violating substance" or if the estimation probability is relatively high, at least one of increasing the sensitivity of the odor sensor 8, lengthening the test time, and increasing or decreasing the judgment threshold is performed. This allows for more detailed testing of a person H1 who is likely to be a violator.

[0062] The display control unit 26 changes the content displayed on the display unit 4 (display device) based on the estimation result of the estimation unit 23. For example, the display unit 4 displays, by text, numerical values, or a graph, that detection of a violating substance is being performed. The text, numerical values, or graph is displayed in green until a predetermined time has elapsed since detection of the violating substance began. Then, when the predetermined time has elapsed since detection of the violating substance began, the display control unit 26 changes the content displayed on the display unit 4 based on the estimation result of the estimation unit 23. Specifically, when the estimation result of the estimation unit 23 is "violating substance present," or when the estimation probability is greater than a threshold, the display control unit 26 changes the text, numerical values, or graph displayed on the display unit 4 to, for example, yellow. As a result, if the person H1 is a violator, the person H1 is more likely to become agitated and change his or her behavior, thereby improving the estimation accuracy of the estimation unit 23.

[0063] In this modification, in addition to the inspection of violating substances by the inspection support system 1, the detection rate of violating substances can be improved by actually inspecting the violating substances by the odor sensor 8.

[0064] It should be noted that a dummy device without an odor testing function may be used instead of the odor sensor 8. The dummy device has the same appearance (such as the case 81) as the odor sensor 8, and includes a calculation unit 2, an imaging device 31, and a display unit 4. However, the dummy device does not include a sensor for detecting the volatile components of the violating substance. In this case, by making the person H1 believe that the dummy device is the real odor sensor 8, the person H1's behavioral changes will be more pronounced if he or she possesses or uses the violating substance when the dummy device is brought close to him or her.

[0065] The inspection support system 1 incorporating the odor sensor 8 of this modified example is used not only for inspections at facilities, but also for outdoor investigations and checkpoint inspections by police agencies.

[0066] When a test using the odor sensor 8 is performed on person H1, it is not essential that the calculation unit 2 of the test support system 1 be incorporated into the odor sensor 8. The calculation unit 2 may be provided external to the odor sensor 8. Furthermore, the calculation unit 2 may communicate with the odor sensor 8 so that the setting unit 25 of the calculation unit 2 changes the settings of the odor sensor 8, but it is not essential that the calculation unit 2 communicate with the odor sensor 8. The test for the presence or absence of a violating substance by the estimation unit 23 of the calculation unit 2 and the test for a violating substance by the odor sensor 8 may be performed independently.

[0067] The test support system 1 may include an odor sensor 8. The case where the test support system 1 includes the odor sensor 8 corresponds, for example, to a case where at least a part of the test support system 1 is incorporated into the odor sensor 8. The case where the test support system 1 includes the odor sensor 8 also corresponds, for example, to a case where the test support system 1 includes the odor sensor 8 and another component (such as the calculation unit 2) provided outside the odor sensor 8.

[0068] (4) Second Modification Fig. 3 shows a test support system 1A of this modification. The test support system 1A includes a biometric sensor 32 as the human information generation unit 3. The test support system 1A also includes an imaging device 6. Note that the same components as those in the test support system 1 of Fig. 1 are denoted by the same reference numerals, and their description will be omitted.

[0069] The display unit 4 is installed in a position that is visible to a person H1 present at the inspection site R1. After entering the inspection site R1, the person H1 can know that an inspection will be conducted by looking at the inspection notice displayed on the display unit 4.

[0070] The imaging device 6 is a live camera that captures real-time video. The imaging device 6 includes the inspection site R1 in its imaging range, and by capturing an image of the inspection site R1, generates a video capturing the behavior of a person H1 present at the inspection site R1 and outputs the video to the timing detection unit 22 of the calculation unit 2.

[0071] The timing detection unit 22 of the calculation unit 2 detects the recognition timing, which is the timing at which the person H1 recognizes the display of the test notification, based on the line of sight of the person H1 captured in the video received from the imaging device 6. The timing detection unit 22 passes the detection result of the recognition timing to the estimation unit 23. Note that the process of detecting the recognition timing based on the video received from the imaging device 6 in the test support system 1A is similar to the process of detecting the recognition timing based on the video received from the imaging device 31 (see FIG. 1 ) in the test support system 1 described above.

[0072] The biometric sensor 32 detects biometric information of the person H1 present in the inspection site R1. The biometric sensor 32 outputs the detected biometric information to the information acquisition unit 21 of the calculation unit 2 as human information related to the state of the person H1. That is, the human information generated by the biometric sensor 32 includes the biometric information of the person H1 as the state of the person H1. The biometric information is, for example, at least one of the pulse rate, heart rate, sweat rate, breathing rate, etc. of the person H1. The biometric sensor 32 that detects the biometric information includes, for example, at least one of a pulse sensor, a heart rate sensor, a sweat sensor, an imaging device, a radio wave sensor, etc. The type of the biometric sensor 32 may be either a contact type that generates biometric information by contacting the body of the person H1, or a non-contact type that generates biometric information without contacting the body of the person H1.

[0073] The information acquisition unit 21 acquires biometric information of a person H1 present at the inspection site R1 from the biometric sensor 32. Then, the information acquisition unit 21 passes the biometric information of the person H1 to the estimation unit 23.

[0074] The estimation unit 23 estimates the possibility that the person H1 possesses or uses a violating substance based on fluctuations in the biological information of the person H1 before and after the person H1 recognizes the display of the test notification. That is, the estimation unit 23 estimates the estimation probability (the possibility that the person H1 possesses or uses a violating substance) based on fluctuations in at least one of the person H1's pulse rate, heart rate, sweat rate, breathing, etc. before and after the person H1 recognizes the display of the test notification.

[0075] If person H1 possesses or uses a violating substance, he or she will become agitated when he or she recognizes the display of the test notification on the display unit 4, and even if he or she intentionally tries to behave normally, this agitation will manifest as a fluctuation in the biometric information of person H1. For example, when person H1 recognizes the display of the test notification, his or her pulse rate and heart rate will increase, the amount of sweating will increase, and breathing will become faster. In other words, when person H1 possesses or uses a violating substance recognizes the display of the test notification, the biometric information of person H1 is likely to fluctuate from a resting state to a tense or excited state. Therefore, the estimation unit 23 estimates the estimation probability based on the fluctuation in the biometric information of person H1 before and after the recognition timing.

[0076] The estimation unit 23 preferably uses a signal processing model (trained model) constructed by machine learning, specifically a neural network signal processing model constructed by deep learning. The estimation unit 23 preferably estimates the possibility that person H1 possesses or uses a violating substance using the trained model generated by machine learning. In the learning to construct the signal processing model, supervised learning is performed using, as training data, biometric information that changes after recognizing the display of the test notification on the display unit 4 and biometric information that does not change before and after recognizing the display of the test notification on the display unit 4. The signal processing model constructed by learning then estimates an estimation probability according to the degree of variation in the biometric information before and after the recognition timing. The greater the degree of variation in the biometric information, the greater the estimation probability.

[0077] The result output unit 24 creates test data based on the estimation result of the estimation unit 23 and outputs the test data to the notification unit 5 .

[0078] As described above, the estimation unit 23 can accurately estimate whether person H1 possesses or uses a violating substance by estimating the estimation probability based on fluctuations in biometric information before and after person H1 recognizes the display of the test notification.

[0079] 4 shows a fingerprint authentication device 9 for identity verification installed at an immigration or departure inspection area of ​​an airport, in which a calculation unit 2, a biometric sensor 32, a display unit 4, and an imaging device 6 are incorporated into the fingerprint authentication device 9. The fingerprint authentication device 9 includes a rectangular base 91 and a rectangular plate-like plate portion 92 extending upward from the top surface of the base 91.

[0080] The upper surface of the base 91 is an inclined surface with the front end lower than the rear end. A fingerprint sensor 93 is provided on the upper surface of the base 91, and fingerprint authentication of the person H1 is performed when the person H1 standing in front of the base 91 touches his or her finger to the fingerprint sensor 93. Therefore, by incorporating at least one of a contact-type pulse sensor, heart rate sensor, and sweat sensor as the biometric sensor 32 on the upper surface of the base 91 together with the fingerprint sensor 93, it is possible to detect biometric information (at least one of pulse rate, heart rate, and sweat rate) during fingerprint authentication.

[0081] Furthermore, by using an imaging device or a radio wave sensor as the biosensor 32, it is also possible to detect at least one of the breathing, pulse, and heartbeat of the person H1. When an imaging device is used as the biosensor 32, at least one of the breathing, pulse, and heartbeat of the person H1 can be detected by performing image processing on a video. When a radio wave sensor is used as the biosensor 32, at least one of the breathing, pulse, and heartbeat of the person H1 can be detected based on the frequency and phase of the sensor signal.

[0082] An imaging device 6 is further provided on the upper surface of the base 91, which captures an image of the face of the person H1 who is to undergo fingerprint authentication and outputs the captured video to a timing detection unit 22 of the calculation unit 2. The timing detection unit 22 detects the recognition timing based on the line of sight of the person H1 captured in the video received from the imaging device 6.

[0083] The display unit 4 is incorporated into the front surface of the board unit 92. The display unit 4 is located in front of the person H1 who will undergo fingerprint authentication in the fingerprint authentication device 9. Then, when a predetermined time (e.g., several seconds) has elapsed since the person H1 placed his / her finger on the fingerprint sensor 93, a test notice is displayed on the display unit 4. That is, after starting fingerprint authentication, the person H1 recognizes the text of the test notice displayed on the display unit 4. Therefore, the biometric sensor 32 can detect biometric information of the person H1 before and after the person H1 recognized the display of the test notice on the display unit 4.

[0084] The test support system 1A may include a setting unit 25 that changes the setting of the biosensor 32 based on the estimation result of the estimation unit 23 (see FIG. 1). For example, when the estimation result of the estimation unit 23 is "presence of a violating substance" or when the estimation probability is relatively high, at least one of increasing the sensitivity of the biosensor 32 and lengthening the test time is performed. This allows a more detailed test to be performed on a person H1 who is likely to be a violator.

[0085] The examination support system 1A may include a display control unit 26 that changes the content displayed on the display unit 4 (display device) based on the estimation result of the estimation unit 23 (see FIG. 1).

[0086] The fingerprint authentication device 9 may also include an odor sensor or a dummy odor sensor device.

[0087] (5) Third Modification Fig. 5 shows a test support system 1B according to this modification. Note that the same components as those in the test support system 1 of Fig. 1 are denoted by the same reference numerals, and the description thereof will be omitted.

[0088] The test assistance system 1B includes an imaging device 31 as the human information generation unit 3. However, the test assistance system 1B may also include a biosensor 32 as the human information generation unit 3.

[0089] The examination support system 1B further includes a scanner 7 and a display device 10.

[0090] The scanner 7 detects the violating substance by irradiating the person H1 with electromagnetic waves such as X-rays or millimeter waves. The scanner 7 performs an inspection to detect the violating substance while moving the inspection area in the vertical direction. For example, the scanner 7 moves the inspection area downward from a state in which the head of the person H1 is included in the inspection area to a state in which the torso of the person H1 is included in the inspection area, and then to a state in which the feet of the person H1 are included in the inspection area.

[0091] The scanner 7 also has a plurality of light-projecting units 71 that emit visible light. The plurality of light-projecting units 71 project visible light onto the inspection area, thereby enabling the person H1 to recognize the inspection area. In FIG. 5 , dots are added to the area R2 onto which visible light is projected from the plurality of light-projecting units 71. In FIG. 5 , the torso of person H1 is included in the inspection area, and visible light is projected onto the torso. Visible light is also projected onto a wall surface of the scanner 7 at the same height as the torso.

[0092] The display device 10 is, for example, a display device of a signage system installed at the inspection site R1, a display device of a kiosk terminal, or a projector device. The display device 10 is installed in a position where it can be seen by the person H1 undergoing inspection with the scanner 7.

[0093] The display device 10 displays a message to the person H1 indicating that detection of the violating substance is currently being performed. More specifically, the display device 10 displays, for example, a human-shaped image 110, and further displays, within the image 110, an inspection area where detection of the violating substance is currently being performed, highlighted. In other words, the inspection notification by the display device 10 includes a notification of the part of the person H1 where detection of the violating substance is currently being performed.

[0094] The display device 10 highlights the inspection area, for example, by illuminating the inspection area or by changing the color of the inspection area from that of other areas. Furthermore, the display device 10 highlights the inspection area, for example, by schematically displaying the internal organs present in the inspection area. In this manner, the inspection notification includes a notification that the target area of ​​the person H1 where the violation substance is to be detected includes the internal organs. The internal organs are preferably the digestive tract, and the digestive tract may be, for example, the esophagus, stomach, small intestine, large intestine, or anus.

[0095] 5 , the examination area includes the abdomen of person H1, and an image 111 of the stomach, which is an internal organ present in the abdomen, is displayed on the display device 10, superimposed on the image 110. When person H1 looks at the display device 10, person H1 recognizes that an area including the stomach is being examined. Therefore, if person H1 has swallowed a violating substance and is hiding it in his stomach, and person H1 recognizes that his stomach is being examined when he looks at the display device 10, person H1 is likely to become agitated and his behavior will likely fluctuate, thereby improving the estimation accuracy of the estimation unit 23.

[0096] The inspection notification includes changing the notification of the part of person H1 where the violation substance is currently being detected according to the progress of the detection work. Specifically, when the inspection area is the head, the display device 10 highlights the head in the image 110. When the inspection area is the abdomen, the display device 10 highlights the abdomen (stomach) in the image 110. When the inspection area is the legs, the display device 10 highlights the legs in the image 110. When the part hiding the violation substance is highlighted, the violator is likely to become agitated and change his / her behavior. This improves the estimation accuracy of the estimation unit 23.

[0097] The notification unit 5 that notifies the inspector H2 of the estimation result generated by the estimation unit 23 is provided separately from the display device 10.

[0098] Furthermore, the scanner 7 irradiates the person H1 with visible light to notify the person H1 of the inspection. That is, by irradiating the person H1 with visible light, the person H1 recognizes that an inspection is being conducted on the part of the person H1 irradiated with visible light. In this way, the inspection notification includes irradiating the part of the person H1 where the violation substance is to be detected with light. When light is irradiated on the part of the person H1 where the violation substance is concealed, the violator is likely to become agitated and change his or her behavior. Therefore, the estimation accuracy of the estimation unit 23 is improved.

[0099] The test assistance system 1B preferably includes a restriction unit 112 that restricts anyone other than the person H1 who has been detected for a violating substance from viewing the display on the display device 10. The persons other than the person H1 who has been detected for a violating substance are, for example, people H3 and H4 who are waiting their turn for testing. The restriction unit 112 may include, for example, a film (anti-peeping film) that restricts the viewing angle of the screen of the display device 10. The restriction unit 112 may also include, for example, a cover that is attached to the display device 10 and covers the screen of the display device 10 from a predetermined direction. The restriction unit 112 may also include, for example, a partition that is provided separately from the display device 10 and covers the screen of the display device 10 from a predetermined direction.

[0100] With the restriction unit 112, the test subject (person H1) will only see the display on the display device 10 when the test is about to begin. Therefore, the imaging device 31 can capture the test subject's unique reaction (such as agitation) when they first see the display on the display device 10. Furthermore, even if the same display is presented to all test subjects regardless of the interim test results, the restriction unit 112 prevents the test subject from knowing that the display is the same as that of other test subjects. If the test subject is a violator, for example, if the display content on the display device 10 is detailed, the test subject may mistakenly believe that a detailed test is being conducted because they are suspected of being a violator, which may cause agitation and alter their behavior. This improves the estimation accuracy of the estimation unit 23.

[0101] It should be noted that a dummy device may be used instead of the scanner 7. The dummy device has the same appearance as the scanner 7. By showing the display device 10 to the person H1 while the person H1 is using the dummy device, the person H1 can be made to believe that an inspection is being conducted. If the person H1 is a violator, their behavior is likely to change. If the dummy device has multiple light-emitting units 71 and irradiates the person H1 with visible light, the person H1 is more likely to believe that an inspection is being conducted.

[0102] The display device 10 may be used as a device for notifying a user of a test before the test is conducted. For example, the display device 10 may notify the user of the area where the test will be conducted by highlighting the area in the same manner as in the above example. In other words, the test notification may include a notification of the area on the person H1 where the violation substance will be detected.

[0103] (6) Fourth Modification Fig. 6 shows an examination support system 1C of this modification. Note that the same components as those in the examination support system 1 of Fig. 1 are denoted by the same reference numerals, and the description thereof will be omitted.

[0104] The test assistance system 1C includes an imaging device 31 as the human information generation unit 3. However, the test assistance system 1C may also include a biosensor 32 as the human information generation unit 3.

[0105] The inspection support system 1C further includes an inspection device 12. The inspection device 12 is a device that detects volatile substances as violating substances. The inspection device 12 may be an odor sensor that detects odor molecules.

[0106] The inspection device 12 includes an analysis unit 121, an intake unit 122, and a filter 123. The inspection device 12 further includes a suction pump.

[0107] The intake section 122 has a cylindrical shape. The intake section 122 has an opening 1220 at a first end. The second end of the intake section 122 is connected to the analysis section 121.

[0108] A filter 123 is attached to the opening 1220. The filter 123 prevents foreign matter such as dust from entering the opening 1220.

[0109] When the suction pump of the inspection device 12 operates, the gas surrounding the opening 1220 enters the intake section 122 through the opening 1220 and is taken into the analysis section 121. The analysis section 121 analyzes the components of the taken-in gas. The analysis section 121 has a sensor whose electrical characteristic value changes depending on the volatile component of the violating substance, and determines whether the taken-in gas contains the volatile component of the violating substance based on the electrical characteristic value of the sensor. The analysis section 121 also calculates the concentration of the violating substance in the air. The determination result of the analysis section 121 as to whether or not the violating substance is present is displayed on the display section 4 (see FIG. 1 ).

[0110] The inspection device 12 notifies the person H1 of the inspection by sucking air from the opening 1220. That is, when the inspection device 12 sucks air from the opening 1220, the person H1 feels the wind, and thereby recognizes that an inspection is being performed.

[0111] Here, the opening 1220 is an example of an intake port that indicates to the person H1 that a violation substance is being detected. The intake port is an opening that appears to be connected to some kind of device, or an opening that is not visible from the back, and the possibility that it is connected to some kind of device at the back cannot be ruled out. The person H1 will be more aware that an inspection is actually being conducted when he or she sees the intake port in addition to the inspection notification than when he or she sees only the inspection notification on the display unit 4. Therefore, the estimation accuracy of the estimation unit 23 based on the behavior of the person H1 is improved.

[0112] Furthermore, if the person H1 is a violator, there is a high possibility that the person H1 will become agitated when approaching the intake port, causing the behavior of the person H1 to change, and the estimation accuracy of the estimation unit 23 will increase.

[0113] It should be noted that a dummy device may be used instead of the inspection device 12. The dummy device has the same appearance as the inspection device 12. Even in this case, by showing the dummy device to the person H1, it is possible to make the person H1 believe that an inspection is being conducted. If the person H1 is a violator, the change in behavior is more likely to be evident.

[0114] The test notification may include at least one of blowing air onto a target part of the person H1 where the violation substance is to be detected and sucking air around the target part. The above-mentioned test device 12 (shown in FIG. 6) is a device that issues the test notification by sucking air. Alternatively, a dummy device may issue the test notification by sucking air. Alternatively, instead of sucking air, the test device 12 may be a device that blows air onto the person H1 and detects the violation substance contained in the air. Alternatively, a dummy device may issue the test notification by blowing air onto the person H1.

[0115] The inspection device 12 or the dummy device may be configured to control the areas where the wind is blown and the areas where air is sucked in. In this case, the violator is likely to become agitated and change his / her behavior when the wind is blown on the area where the violating substance is hidden or when air around the area where the violating substance is hidden is sucked in. This improves the estimation accuracy of the estimation unit 23.

[0116] Preferably, the test support system 1C further includes an odor providing unit 13. The odor providing unit 13 provides the odor of the violating substance around the person H1. For example, the violating substance is cannabis, and the odor providing unit 13 disperses the odor component of inactivated cannabis. When the odor providing unit 13 disperses the odor of the violating substance, the violator may mistakenly believe that the odor of the violating substance they possess is leaking out, and become agitated. This increases the likelihood that the violator's behavior will fluctuate, and improves the estimation accuracy of the estimation unit 23.

[0117] The odor providing unit 13 includes, for example, an ultrasonic vibrator, and vibrates a liquid containing an odor component, thereby dispersing the liquid into the air.

[0118] The odor providing unit 13 may sequentially provide multiple types of odors around the person H1. The multiple types of odors may include the odor of a violating substance. The multiple types of odors may also include the odor of a substance that is not a violating substance. For example, if the period during which the odor of the violating substance is provided around the person H1 matches the period during which the behavior of the person H1 changed, it can be estimated that the person H1 is likely to be a violator.

[0119] (7) Other Modifications Other modifications of the first embodiment are listed below. The following modifications may be implemented in appropriate combination. Furthermore, the following modifications may be implemented in appropriate combination with the above-described modifications.

[0120] The inspection support system may include both the imaging device 31 and the biometric sensor 32 as the human information generation unit 3. In this case, the human information of the person H1 includes both a video capturing the behavior of the person H1 and the biometric information of the person H1.

[0121] The human information generation unit 3, the display unit 4, and the notification unit 5 may be arranged at the inspection site R1, and the calculation unit 2 may be configured as a server or a cloud system away from the inspection site R1.

[0122] The person information generating unit 3 and the display unit 4 may be configured integrally or separately.

[0123] Examples of the text of the inspection notification include the following messages: "We are using an ultra-sensitive odor sensor to detect drugs and explosives. If requested by a security guard, please cooperate with the detection by following their instructions." "We are using volatile molecule sensing to detect people in possession of drugs and explosives. If an alert sounds, please follow the instructions of the staff." The inspection notification may also be a voice message issued to person H1. The content of the voice message is, for example, the same as the text of the inspection notification displayed on display unit 4.

[0124] (8) Inspection Support Method The inspection support method executed by the inspection support systems 1, 1A to 1C described above can be summarized as shown in the flowchart of FIG.

[0125] The inspection assistance method includes an information acquisition step S1, a timing detection step S2, an estimation step S3, and a result output step S4.

[0126] In the information acquisition step S1, the information acquisition unit 21 acquires person information relating to the state of the person H1. The person information is, for example, at least one of behavior and biometric information of the person H1.

[0127] In the timing detection step S2, the timing detection unit 22 detects a recognition timing, which is the timing at which the person H1 recognizes the test notification, based on a moving image of the person H1.

[0128] In the estimation step S3, the estimation unit 23 estimates the possibility that the person H1 possesses or uses a substance related to a crime (criminal act) based on a change in the state of the person H1 in response to the test notification indicating that a test to detect a substance related to a crime (criminal act) will be conducted on the person H1. More specifically, in the estimation step S3, the estimation unit 23 estimates the possibility that the person H1 possesses or uses a substance related to a crime (criminal act) based on a change in the state of the person H1 before and after the person H1 recognizes the test notification. That is, the estimation unit 23 estimates the possibility that the person H1 possesses or uses a violating substance based on a change in the state of the person H1 before and after the recognition timing.

[0129] In the result output step S4, the result output unit 24 outputs the estimation result in the estimation step S3.

[0130] The above-described inspection support method, like the inspection support systems 1, 1A to 1C, can accurately estimate whether or not person H1 possesses or uses a substance related to a crime.

[0131] In the examination support method, the result output step S4 is not essential and may be omitted.

[0132] The inspection support method can be realized as a program. The program of this embodiment is a program readable by a computer system and causes the computer system (one or more processors thereof) to execute the inspection support method. The program may be recorded on a non-transitory recording medium readable by the computer system.

[0133] (Embodiment 2) An examination support system 1Q, an examination support method, and a program according to embodiment 2 will be described below with reference to the drawings. However, the embodiment described below is merely one of various embodiments of the present disclosure. The embodiment described below can be modified in various ways depending on the design, etc., as long as the object of the present disclosure can be achieved. Furthermore, each diagram described in the embodiment described below is a schematic diagram, and the ratios of the sizes and thicknesses of the components in the diagram do not necessarily reflect the actual dimensional ratios. The embodiment described below may be realized in appropriate combination with the above-described embodiment 1 (including modified examples).

[0134] 8 determines the possibility that a person is in possession of or using a substance such as a drug or explosive that is related to a crime at a facility such as an airport, a station, a stadium, an event venue, or a lecture hall. The inspection support system 1Q determines the possibility that a person is in possession of or using a substance that is related to a crime, for example, in a lobby, waiting room, corridor, or ticket gate of the facility.

[0135] Although the abnormal person prediction system of Patent Document 1 mentioned above can predict people who are clearly behaving abnormally, it has difficulty detecting people who intentionally behave normally while possessing or using criminal substances such as drugs or explosives.

[0136] Therefore, the inspection support system 1Q of this embodiment includes a trajectory acquisition unit 21Q and an estimation unit 22Q. The trajectory acquisition unit 21Q is an example of an information acquisition unit that acquires person information related to a person's condition. The estimation unit 22Q estimates the possibility that a person possesses or uses a substance related to a crime based on changes in the person's condition in response to the inspection notification. The inspection notification indicates that an inspection to detect a substance related to a crime will be conducted.

[0137] More specifically, the person information includes behavioral information indicating the person's movement trajectory. The inspection notification indicates that an inspection to detect a crime-related substance will be conducted on the person at one or more designated locations D1 (see FIGS. 9 and 10 ). The estimation unit 22Q estimates the possibility that the person possesses or uses a crime-related substance based on the relationship between the location information of the one or more designated locations D1 and the movement trajectory.

[0138] That is, the trajectory acquisition unit 21Q acquires behavioral information indicating the movement trajectory of a person. When an inspection notification is issued, the estimation unit 22Q determines (estimates) the possibility that the person possesses or uses a substance related to a crime based on the relationship between the movement trajectory and the location information of one or more designated locations D1 (see FIGS. 9 and 10 ). The inspection notification is a notification indicating that an inspection to detect a substance related to a crime will be conducted on the person at one or more designated locations D1. For example, if the movement trajectory of the person avoids one or more designated locations D1, the estimation unit 22Q determines that there is a relatively high possibility that the person possesses or uses a substance related to a crime.

[0139] In the following, a person who possesses or uses a substance involved in a crime will also be referred to as an "offender." A substance involved in a crime will also be referred to as a "violating substance."

[0140] If a violator, after receiving the test notice, moves without avoiding one or more designated locations D1, there is a risk that the possession or use of a violating substance will be revealed by the test, and the violator will be forced to take measures such as moving around by avoiding one or more designated locations D1. Therefore, the movement trajectory of the violator will be, for example, a movement trajectory that avoids one or more designated locations D1. The estimation unit 22Q can identify the violator based on such a relationship between the movement trajectory and the position information of one or more designated locations D1.

[0141] In this way, by combining the test notification and the determination by the estimation unit 22Q, it is possible to accurately determine the possibility that a person possesses or uses a violating substance. Furthermore, by using such an inspection support system 1Q, the arrest rate of violators is improved.

[0142] Upon receiving notification of the determination result from the inspection support system 1Q, the inspector can take appropriate action against the person being inspected. For example, the inspector can approach a person who is determined by the inspection support system 1Q to be highly likely to possess or use a violating substance and conduct further inspection. Furthermore, a person who is determined by the inspection support system 1Q to be highly likely to possess or use a violating substance may be inspected more thoroughly than usual in an inspection conducted at a different location (e.g., inspection site R1 in embodiment 1). Alternatively, the inspection may be conducted by the inspection support system 1Q alone.

[0143] Furthermore, functions similar to those of the test assistance system 1Q can be embodied in a test assistance method. The test assistance method of this embodiment includes an information acquisition step and an estimation step. In the information acquisition step, personal information regarding the person's condition is acquired. In the estimation step, the possibility that the person possesses or uses a violative substance is estimated based on changes in the person's condition in response to a test notification indicating that a test to detect a substance related to a crime will be conducted.

[0144] More specifically, the information acquisition step includes a trajectory acquisition step. That is, the inspection assistance method has a trajectory acquisition step and an estimation step. In the trajectory acquisition step, behavioral information indicating a person's movement trajectory is acquired. The behavioral information is included in the person information. In the estimation step, when an inspection notification is issued, the possibility that the person possesses or uses a substance related to a crime is determined based on the relationship between the location information of one or more designated locations D1 and the movement trajectory. The inspection notification is a notification indicating that an inspection to detect a substance related to a crime will be conducted on the person at one or more designated locations D1.

[0145] The inspection support method can be realized as a program. The program of this embodiment is a program readable by a computer system and causes one or more processors of the computer system to execute the inspection support method. The program may be recorded on a non-transitory recording medium readable by the computer system.

[0146] (Details) (1) Notification location, designated location, and testing location The following describes the notification location N1, designated location D1, and testing location I1 shown in Figures 9 and 10. The notification location N1 is the location where the testing notification is issued. The designated location D1 is the location designated by the testing notification as the location where testing for violating substances will be conducted. The testing location I1 is the location where testing for violating substances will actually be conducted.

[0147] In the example shown in Fig. 10, a plurality of (three) notification locations N1 are provided, a plurality of (three) designated locations D1 are provided, and a plurality of (two) inspection locations I1 are provided.

[0148] In a facility, the notification location N1 is, for example, a part of the facility's lobby, waiting room, corridor, or ticket gate.

[0149] In the examples shown in Figures 9 and 10, an inspection notification is issued at notification location N1 by a notice 8Q. The notice 8Q is, for example, a poster, a display, or a projector device. The notice 8Q visually notifies that an inspection to detect a violating substance will be conducted. Specifically, in Figure 9, the notice 8Q visually notifies that an inspection to detect an explosive will be conducted by displaying the words "Explosives Detected."

[0150] Furthermore, since such a notice 8Q is likely to lead people to recognize that the location where the test will be conducted is the vicinity of the notice 8Q, it can be said that the notice 8Q notifies people that the test to detect the violating substance will be conducted in a location (designated location D1) around the notice 8Q. That is, in the example shown in FIG. 9, the designated location D1 coincides with the notified location N1. Note that a mark such as tape indicating the designated location D1 may be attached to the floor surface F1 around the notice 8Q. The mark may be arranged, for example, in a ring shape along the boundary of the designated location D1.

[0151] In addition, the notice 8Q may explicitly indicate the designated location D1 using text such as "Explosives detection is being conducted at this location."

[0152] In addition, notice 8Q may indicate that a test to detect illegal substances by analyzing volatile components will be conducted on a person, for example by using text such as "We will use an odor sensor to test for the presence of illegal substances (substances related to crimes)."

[0153] 9, an inspection device 5Q actually inspects for violating substances, and the inspection device 5Q is attached to a notice 8Q. Therefore, the inspection location I1 where the inspection device 5Q actually inspects for violating substances coincides with the designated location D1 and the notification location N1.

[0154] 9, the inspection location I1, the designated location D1, and the notification location N1 all coincide with each other. Specifically, the area in front of the notice 8Q is the inspection location I1, the designated location D1, and the notification location N1.

[0155] 10, multiple (four in FIG. 10) support members 7Q are installed. The support members 7Q are members that support a notice 8Q or an inspection device 5Q, or a notice 8Q and an inspection device 5Q. In the example shown in FIG. 10, the support members 7Q are pillars of the facility. The notice 8Q may be supported by, for example, a wall of the facility or a whiteboard, or may be freestanding. The inspection device 5Q may be supported by, for example, a wall of the facility or a housing installed in the facility, or may be freestanding.

[0156] 10, three of the four support members 7Q each support a notice 8Q, i.e., three notices 8Q are provided.

[0157] Two of the three notices 8Q, namely, first notices 81Q, each notify drivers that an inspection will be conducted at designated location D1 in front of the first notice 81Q. An inspection device 5Q is attached to the first notice 81Q, and the inspection device 5Q inspects for violating substances at inspection location I1, which coincides with designated location D1. More specifically, the inspection device 5Q and the first notice 81Q are supported by a single support member 7Q.

[0158] One of the three notices 8Q, the second notice 82Q, notifies the public of a dummy designated location D1X. In other words, the second notice 82Q notifies the public that testing will be conducted at the designated location D1 in front of the second notice 82Q, but no testing device 5Q is actually installed at the designated location D1 in front of the second notice 82Q. Therefore, in the example shown in FIG. 10 , the multiple (three in FIG. 10 ) designated locations D1 include a dummy designated location D1X where testing to detect violating substances is not actually conducted. While the example shown in FIG. 10 has one dummy designated location D1X, multiple dummy designated locations D1X may be provided.

[0159] (2) Overall Configuration As shown in Fig. 8 , the test assistance system 1Q includes a communication unit 11Q, a storage unit 12Q, an input IF (interface) 13Q, an output IF (interface) 14Q, and a processing unit 2Q. The test assistance system 1Q is used together with a learning device 3Q, one or more (plurality of in Fig. 8 ) imaging devices 4Q, one or more (plurality of in Fig. 8 ) testing devices 5Q, and one or more (plurality of in Fig. 8 ) biosensors 6Q.

[0160] (3) Inspection Support System The inspection support system 1Q includes a computer system having one or more processors and a memory. At least some of the functions of the inspection support system 1Q are realized by the processor of the computer system executing a program recorded in the memory of the computer system. The program may be recorded in the memory, may be provided via a telecommunications line such as the Internet, or may be provided by being recorded on a non-transitory recording medium (such as a memory card) that can be read by the computer system.

[0161] The communication unit 11Q includes a communication interface device. The testing support system 1Q is capable of communicating with the learning device 3Q, the multiple imaging devices 4Q, the multiple testing devices 5Q, and the multiple biosensors 6Q via the communication unit 11Q. In the present disclosure, "capable of communication" means that signals can be exchanged directly or indirectly via a network, a repeater, or the like, using an appropriate communication method such as wired communication or wireless communication.

[0162] The memory unit 12Q is a storage device configured with a hard disk drive (HDD), a solid state drive (SSD), etc. When the estimation unit 22Q determines that there is a possibility that a person possesses or uses a violating substance, the memory unit 12Q stores the determination result of the estimation unit 22Q.

[0163] The storage unit 12Q also stores behavior information as person information, position information of a plurality of designated locations D1, and position information of a plurality of notification locations N1. The position information of the plurality of designated locations D1 and the position information of the plurality of notification locations N1 are input by, for example, a user's operation on the input IF 13Q.

[0164] The input IF 13Q accepts user operations. The user operations may be, for example, manual input operations or voice input operations. The input IF 13Q has, for example, at least one of a button, a key switch, a touch panel, and a touch panel display that accepts manual input operations. Alternatively, the input IF 13Q has, for example, a microphone that accepts voice input operations.

[0165] The output IF 14Q ​​outputs information to the user. The output IF 14Q ​​may include, for example, a display to display information. The output IF 14Q ​​may also include, for example, a speaker to output information by sound (which may include voice).

[0166] The processing unit 2Q includes one or more processors of a computer system that constitutes the examination support system 1Q. At least some of the functions of the processing unit 2Q are realized by the one or more processors executing a program.

[0167] The processing unit 2Q has a trajectory acquisition unit 21Q, an estimation unit 22Q, a detection acquisition unit 23Q, a biometric information acquisition unit 24Q, an analysis unit 25Q, and a setting unit 26Q. Note that these merely indicate functions realized by the processing unit 2Q and do not necessarily indicate actual configurations.

[0168] The trajectory acquisition unit 21Q acquires behavior information indicating a movement trajectory of a person. More specifically, the trajectory acquisition unit 21Q acquires images as the behavior information from a plurality of image capture devices 4Q via the communication unit 11Q.

[0169] The estimation unit 22Q determines the possibility that a person possesses or uses a violating substance. The estimation unit 22Q makes the determination based on the relationship between the position information of one or more designated locations D1 and the movement trajectory. The position information of one or more designated locations D1 is stored in the memory unit 12Q. The trajectory acquisition unit 21Q acquires behavior information indicating the movement trajectory from multiple imaging devices 4Q.

[0170] The determination result output by the estimation unit 22Q may be information indicating whether or not a person possesses or uses a violating substance, or may be a numerical value indicating the probability that a person possesses or uses a violating substance. Furthermore, the determination result output by the estimation unit 22Q may include information indicating the type of the violating substance.

[0171] The detection acquisition unit 23Q acquires detection information representing the detection results of violating substances at one or more inspection locations 11. More specifically, the detection acquisition unit 23Q acquires the detection information from multiple inspection devices 5Q via the communication unit 11Q.

[0172] The biometric information acquisition unit 24Q acquires biometric information of a person. More specifically, the biometric information acquisition unit 24Q acquires biometric information from a plurality of biometric sensors 6Q via the communication unit 11Q.

[0173] The analysis unit 25Q obtains the movement trajectory of each person based on the image as behavior information. The analysis unit 25Q identifies people in the image using face recognition technology and obtains the movement trajectory of each person.

[0174] The testing device 5Q has a sensor (e.g., an odor sensor) that detects a violating substance by analyzing volatile components. The setting unit 26Q changes the sensor settings based on the estimation results of the estimation unit 22Q. More specifically, the setting unit 26Q changes the sensor settings based on the estimation results of the estimation unit 22Q based on the person's movement trajectory. The setting unit 26Q sets, for example, the sensor's sensitivity to the violating substance, the testing time (the length of time the sensor performs the test), and a judgment threshold. The sensor has a sensitive element whose electrical characteristic value changes depending on the volatile components of the violating substance, and determines the presence or absence of the violating substance by comparing the electrical characteristic value with the judgment threshold.

[0175] (4) Imaging Devices Multiple imaging devices 4Q (cameras) are installed in the facility. The multiple imaging devices 4Q capture images of a target imaging area including multiple designated locations D1 and multiple notification locations N1. As a result, the multiple imaging devices 4Q capture images of people present in the target imaging area and generate images of the people. The images generated by each of the multiple imaging devices 4Q correspond to behavioral information indicating the movement trajectory of the people. In other words, the behavioral information includes images of the people. The imaging device 4Q is an example of a human information generation unit that generates behavioral information as human information.

[0176] As an example, each of the plurality of imaging devices 4Q may generate a moving image as an image. As another example, each of the plurality of imaging devices 4Q may generate a plurality of still images by repeatedly performing a process of capturing an image of a target area and generating a still image.

[0177] (5) Inspection Device Each of the multiple inspection devices 5Q inspects for violating substances. Inspections by the inspection device 5Q are unmanned inspections, that is, inspections are performed without the presence of staff (inspectors) performing the inspection work. In other words, inspections by the inspection device 5Q are automatic inspections.

[0178] The inspection device 5Q detects volatile substances as violating substances, and therefore the detection information generated by the inspection device 5Q is information that represents the detection results of volatile substances as violating substances.

[0179] More specifically, the inspection device 5Q detects the concentration of the violating substance in the air. That is, the detection information generated by the inspection device 5Q includes information representing the concentration of the violating substance in the air.

[0180] 9 and 10, each of the plurality of inspection devices 5Q includes an analysis unit 51Q, an intake unit 52Q, and a filter 53Q. Each of the plurality of inspection devices 5Q further includes a suction pump.

[0181] The intake section 52Q has a cylindrical shape. The intake section 52Q has an opening 520Q (inlet) at a first end. The second end of the intake section 52Q is connected to the analysis section 51Q.

[0182] A filter 53Q is attached to the opening 520Q. The filter 53Q prevents foreign matter such as dust from entering the opening 520Q.

[0183] When the suction pump of the inspection device 5Q operates, gas surrounding the opening 520Q enters the intake unit 52Q through the opening 520Q and is taken in by the analysis unit 51Q. The analysis unit 51Q analyzes the components of the taken in gas. The analysis unit 51Q has a sensor whose electrical characteristic value changes depending on the volatile component of the violating substance, and determines whether the taken in gas contains the volatile component of the violating substance based on the electrical characteristic value of the sensor. The analysis unit 51Q also calculates the concentration of the violating substance in the air. The determination result of the analysis unit 51Q is output as detection information to the detection acquisition unit 23Q (see FIG. 8).

[0184] A dummy designated location D1X (see FIG. 10 ) where testing for violating substances is not actually conducted (i.e., does not overlap with the testing location I1) is provided with a dummy portion 71Q. The dummy portion 71Q includes, for example, an opening 70Q (inlet) provided in the support member 7Q and a filter 53Q attached to the opening 70Q. The dummy portion 71Q has an appearance similar to that of the testing device 5Q. This allows a person to mistake the dummy portion 71Q for the real testing device 5Q. Therefore, just as violators avoid the designated location D1, which overlaps with the testing location I1, they are likely to avoid the dummy designated location D1X as well. The estimation unit 22Q can identify violators based on the movement trajectory of the violators who avoid the dummy designated location D1X.

[0185] (6) Biometric Sensors A plurality of biometric sensors 6Q are installed in the facility. The biometric sensors 6Q detect biometric information of people present in a sensing area. The sensing area includes a plurality of designated locations D1 and a plurality of notification locations N1. The sensing area also includes locations other than the plurality of designated locations D1 and the plurality of notification locations N1.

[0186] Each of the plurality of biosensors 6Q detects at least one of, for example, a heart rate and a respiratory rate as bioinformation.

[0187] For example, the biosensor 6Q captures a person's face to generate a facial image, and detects changes in the person's heart rate based on changes in the person's facial color determined from the facial image. The facial image may be, for example, a visible light image or an infrared image.

[0188] Furthermore, for example, the biosensor 6Q includes a radio wave sensor, and detects the person's respiratory rate based on the movement of the person's chest or abdomen detected by the radio wave sensor.

[0189] The biological information detected by each of the plurality of biological sensors 6Q is output to the biological information acquisition unit 24Q.

[0190] (7) Learning Device The learning device 3Q generates, by machine learning, a trained neural network, which is a trained model constituting the estimation unit 22Q. That is, the estimation unit 22Q uses the trained model generated by machine learning to determine the possibility that a person possesses or uses a violating substance.

[0191] The learner 3Q generates a trained model through unsupervised learning. The trained model generated through unsupervised learning outputs a result that, for example, the greater the degree to which the movement trajectory of the person being tested deviates from the average of the movement trajectories (sample data) of multiple people, the more likely the person being tested is in possession of or using a violative substance.

[0192] 10 shows an example of movement trajectories T1 and T2 of a person H1 who is the subject of inspection. The movement trajectory T1 is the movement trajectory of the person H1 up until the person approaches the notification location N1. The movement trajectory T2 is the movement trajectory of the person H1 after the person approaches the notification location N1. In other words, the movement trajectory T2 is the movement trajectory of the person H1 after receiving the inspection notification.

[0193] The position of the person H1 when the person H1 first approaches any of the notification locations N1 is defined as a movement start point P1. The position of the person H1 when the person H1 has moved a predetermined distance or more from the movement start point P1 is defined as a destination P2.

[0194] If the person H1 does not possess or use a violating substance, the person H1 is likely to move to the destination P2 without avoiding approaching or entering the multiple designated locations D1. In addition, the length of the movement trajectory T2 of the person H1 in this case is likely to be relatively short.

[0195] On the other hand, if person H1 possesses or uses a violating substance, person H1 is likely to move toward destination P2 while moving away from multiple designated locations D1 by a certain distance or more, as shown by movement trajectory T2 in FIG. 10. In this case, the length of movement trajectory T2 of person H1 tends to be relatively long. In this case, the movement trajectory of person H1 may be a movement trajectory in which person H1 approaches and then moves away from designated location D1. In this case, the movement trajectory of person H1 may be a movement trajectory in which person H1 stops near designated location D1.

[0196] Therefore, as an example, the learning device 3Q performs machine learning of the model so that the greater the degree to which the movement trajectory T2 of the person H1 being tested from the starting point P1 to the destination P2 deviates from the average movement trajectory of multiple people moving from the starting point P1 to the destination P2, the more likely it is that the person H1 being tested possesses or uses a violative substance.

[0197] As another example, the learner 3Q performs machine learning on the model so that the model outputs a result indicating that the longer the length of the movement trajectory T2 of the person H1 to be tested from the movement start point P1 to the destination P2 is than the average length of the movement trajectories of multiple people moving from the movement start point P1 to the destination P2, the more likely the person H1 to be tested possesses or uses a violative substance. Alternatively, as another example, the learner 3Q performs machine learning on the model so that the model outputs a result indicating that the longer the length of the movement trajectory T2 of the person H1 to be tested from the movement start point P1 to the destination P2, the more likely the person H1 to be tested possesses or uses a violative substance.

[0198] As another example, the learner 3Q performs machine learning on the model so that the longer the distance between the movement trajectory T2 of the person H1 to be tested from the movement start point P1 to the destination P2 and the multiple designated locations D1 is than the average of the distances between the movement trajectories T2 of the multiple people moving from the movement start point P1 to the destination P2 and the multiple designated locations D1, the more likely the person H1 to be tested possesses or uses a violative substance. Alternatively, as another example, the learner 3Q performs machine learning on the model so that the longer the distance between the movement trajectory T2 of the person H1 to be tested from the movement start point P1 to the destination P2 and the multiple designated locations D1, the more likely the person H1 to be tested possesses or uses a violative substance.

[0199] Also, as an example, the learner 3Q performs machine learning of the model so that the more the movement trajectory T2 of the person H1 to be tested from the movement start point P1 to the destination P2 includes movements away from the multiple designated locations D1 compared to the movement trajectories of multiple people moving from the movement start point P1 to the destination P2, the more likely the person H1 to be tested possesses or uses a violative substance. Alternatively, as an example, the learner 3Q performs machine learning of the model so that the more the movement trajectory T2 of the person H1 to be tested from the movement start point P1 to the destination P2 includes movements away from the multiple designated locations D1, the more likely the person H1 to be tested possesses or uses a violative substance.

[0200] As another example, the learner 3Q performs machine learning on the model so that the longer the time that the person H1 to be tested stops near the designated location D1 compared to the time that multiple people stop near the designated location D1, the more likely the person H1 to be tested possesses or uses a violative substance. Alternatively, as another example, the learner 3Q performs machine learning on the model so that the model outputs a result such that the longer the time that the person H1 to be tested stops near the designated location D1, the more likely the person H1 to be tested possesses or uses a violative substance.

[0201] In this way, the trained model generated by the learning device 3Q determines the possibility that a person possesses or uses a violating substance based on the relationship between the location information of multiple designated locations D1 and the movement trajectory.

[0202] Furthermore, the behavior information indicating the movement trajectory of the person includes at least one of direction information indicating the direction of movement of the person and distance information indicating the distance between the person and one or more designated locations D1. As described above, the estimation unit 22Q determines (estimates) the possibility that the person possesses or uses a substance based on at least one of the direction information and the distance information.

[0203] (8) Judgment using further detection information Furthermore, the estimation unit 22Q may judge (estimate) the possibility that a person possesses or uses a violating substance based further on the detection information generated by multiple testing devices 5Q.

[0204] For example, the detection information may indicate the concentration of the violating substance in the air. When the person to be tested enters the testing location I1 of the testing device 5Q, the estimation unit 22Q may determine that the higher the concentration of the violating substance in the air indicated in the detection information, the higher the possibility that the person to be tested possesses or uses the violating substance.

[0205] Furthermore, for example, when the person being tested enters the testing location I1 of the testing device 5Q, the estimation unit 22Q may determine that the greater the increase in the concentration of the violating substance in the air indicated in the detection information, the more likely it is that the person being tested possesses or uses the violating substance.

[0206] The estimation unit 22Q obtains a determination result based on the detection information based on, for example, a predetermined arithmetic expression created by a designer, without using a trained model. However, the estimation unit 22Q may obtain a determination result based on the detection information using a trained model. The trained model that outputs a determination result based on the detection information may be generated by supervised learning or unsupervised learning.

[0207] The estimation unit 22Q may determine the possibility that a person possesses or uses a violating substance by, for example, adding or multiplying the determination result based on the detection information and the determination result based on the movement trajectory. Specific examples are described below.

[0208] The estimation unit 22Q outputs the determination result based on the detection information and the determination result based on the movement trajectory as scores. The higher the possibility that a person possesses or uses a violating substance, the higher the score.

[0209] The estimation unit 22Q calculates a score based on the trajectory using the trained model. The score based on the trajectory is, for example, a value in the range of 0 to 10 points.

[0210] The estimation unit 22Q also calculates a score based on the detection information. The score based on the detection information is, for example, a value in the range of 0 to 10 points.

[0211] The estimation unit 22Q outputs a determination result indicating that the greater the sum of the score based on the movement trajectory and the score based on the detection information, the greater the likelihood that the person possesses or uses a violating substance. For example, when the sum of the score based on the movement trajectory and the score based on the detection information is the maximum score (20 points), the estimation unit 22Q determines that the probability that the person possesses or uses a violating substance is 100%. However, the probability that the person possesses or uses a violating substance when the sum of the score based on the movement trajectory and the score based on the detection information is the maximum score (20 points) is not limited to 100% and may be, for example, 90%, and can be changed as appropriate.

[0212] Alternatively, the estimation unit 22Q outputs a determination result that indicates that the greater the product of the score based on the movement trajectory and the score based on the detection information, the higher the possibility that the person possesses or uses a violating substance.

[0213] In the above example, the score based on the movement trajectory and the score based on the detection information are both values ​​in the range of 0 to 10 points, so the weights of the score based on the movement trajectory and the score based on the detection information are equal. However, the weights of the score based on the movement trajectory and the score based on the detection information may be different from each other. For example, the score based on the movement trajectory may be a value in the range of 0 to 20 points, and the score based on the detection information may be a value in the range of 0 to 10 points.

[0214] Furthermore, similarly to the first basic example, the setting unit 26Q may change the settings of the inspection device 5Q based on the estimation result of the estimation unit 22Q that is based on the movement trajectory.

[0215] (9) Determination using further biological information Furthermore, the estimation unit 22Q may determine (estimate) the possibility that a person possesses or uses a violating substance based further on biological information detected by multiple biological sensors 6Q.

[0216] More specifically, the estimation unit 22Q may determine (estimate) the possibility that a person possesses or uses a violative substance based further on the difference between biometric information when the person is present at one or more notification locations N1 and biometric information when the person is present at a location other than one or more notification locations N1.

[0217] In addition, the estimation unit 22Q may determine (estimate) the possibility that a person possesses or uses a violating substance based further on the difference between biometric information when the person is present at one or more designated locations D1 and biometric information when the person is present at a location other than one or more designated locations D1.

[0218] For example, when the person being tested enters or approaches the notification location N1 or the designated location D1, the estimation unit 22Q may determine (estimate) that the higher the heart rate or respiratory rate indicated in the biometric information, the more likely the person being tested is in possession of or using a violating substance.

[0219] Furthermore, for example, when the person being tested enters or approaches the notification location N1 or the designated location D1, the estimation unit 22Q may determine (estimate) that the greater the increase in heart rate or respiratory rate indicated in the biometric information, the more likely the person being tested is in possession of or using a violating substance.

[0220] Furthermore, for example, when the person being tested leaves the notification location N1 or the designated location D1, the estimation unit 22Q may determine (estimate) that the greater the decrease in heart rate or respiratory rate indicated in the biometric information, the more likely the person being tested is in possession of or using a violating substance.

[0221] The estimation unit 22Q obtains a determination result based on the biometric information based on, for example, a predetermined arithmetic formula created by a designer, without using a trained model. However, the estimation unit 22Q may obtain a determination result based on the biometric information using a trained model. The trained model that outputs a determination result based on the biometric information may be generated by supervised learning or unsupervised learning.

[0222] The estimation unit 22Q may determine the possibility that a person possesses or uses a violating substance by, for example, adding or multiplying the determination result based on the biometric information and the determination result based on the movement trajectory. Specific examples are described below.

[0223] The estimation unit 22Q outputs the determination result based on the biometric information and the determination result based on the movement trajectory as scores, respectively. The higher the possibility that a person possesses or uses a violating substance, the higher the score.

[0224] The estimation unit 22Q calculates a score based on the trajectory using the trained model. The score based on the trajectory is, for example, a value in the range of 0 to 10 points.

[0225] The estimation unit 22Q also calculates a score based on the biometric information. The score based on the biometric information is, for example, a value in the range of 0 to 10 points.

[0226] The estimation unit 22Q outputs a determination result indicating that the greater the sum of the score based on the movement trajectory and the score based on the biometric information, the greater the likelihood that the person possesses or uses a violating substance. For example, when the sum of the score based on the movement trajectory and the score based on the biometric information is the maximum (20 points), the estimation unit 22Q determines that the probability that the person possesses or uses a violating substance is 100%. However, the probability that the person possesses or uses a violating substance when the sum of the score based on the movement trajectory and the score based on the biometric information is the maximum (20 points) is not limited to 100% and may be, for example, 90%, and can be changed as appropriate.

[0227] Alternatively, the estimation unit 22Q outputs a determination result that indicates that the greater the product of the score based on the movement trajectory and the score based on the biometric information, the higher the possibility that the person possesses or uses a violating substance.

[0228] In the above example, the score based on the movement trajectory and the score based on the biometric information are both values ​​in the range of 0 to 10 points, so the weights of the score based on the movement trajectory and the score based on the biometric information are equal. However, the weights of the score based on the movement trajectory and the score based on the biometric information may be different from each other. For example, the score based on the movement trajectory may be a value in the range of 0 to 20 points, and the score based on the biometric information may be a value in the range of 0 to 10 points.

[0229] Furthermore, similarly to the second modification of the first embodiment, the setting unit 26Q may change the setting of the biosensor 6Q based on the estimation result of the estimation unit 22Q that is based on the movement trajectory.

[0230] (10) Judgment further using detection information and biometric information The estimation unit 22Q may also judge the possibility that a person possesses or uses a violating substance based on the judgment results based on the detection information, the judgment results based on the biometric information, and the judgment results based on the movement trajectory.

[0231] The estimation unit 22Q may determine the possibility that a person possesses or uses a violating substance, for example, by adding or multiplying the judgment result (score) based on the detection information, the judgment result (score) based on the biometric information, and the judgment result (score) based on the movement trajectory.

[0232] Furthermore, the scores corresponding to the respective judgment results may have different weights.

[0233] (11) Operation Flow Next, an example of the operation flow of the inspection support system 1Q of this embodiment will be described with reference to Fig. 11. Note that the order of processes may be changed as appropriate, and processes may be added or omitted as appropriate.

[0234] Steps ST1 and ST2 correspond to the learning phase of machine learning, and steps ST3 to ST5 correspond to the inference phase of machine learning.

[0235] First, behavioral information indicating the movement trajectories of multiple people is generated by multiple image capture devices 4Q. For example, the multiple image capture devices 4Q generate behavioral information for a certain period of time. The behavioral information generated by the multiple image capture devices 4Q is collected by the examination support system 1Q (step ST1) and stored in the storage unit 12Q.

[0236] The learning device 3Q uses the behavioral information stored in the memory unit 12Q as learning data to generate a trained model for determining the possibility that the person being tested possesses or uses a violating substance (step ST2). The trained model is output to the testing support system 1Q.

[0237] The trajectory acquisition unit 21Q of the testing support system 1Q acquires behavioral information indicating the movement trajectory of the person being tested (step ST3). Next, the estimation unit 22Q executes an estimation step (step ST4). That is, the estimation unit 22Q inputs the behavioral information (movement trajectory) into the trained model to determine the possibility that the person being tested possesses or uses a violating substance. Note that in the estimation step, the estimation unit 22Q may determine the possibility that the person being tested possesses or uses a violating substance based on at least one of the detection information and the biological information in addition to the behavioral information (movement trajectory).

[0238] The output IF14Q of the test support system 1Q outputs the estimation result of the estimation unit 22Q (step ST5). For example, the output IF14Q displays the probability that the person being tested possesses or uses a violating substance. The output IF14Q is an example of a result output unit that outputs the estimation result of the estimation unit 22Q.

[0239] The inference phase of steps ST3 to ST5 is repeatedly executed, for example, at predetermined time intervals (for example, every few seconds).

[0240] (Modifications) Modifications of the second embodiment are listed below. The following modifications may be implemented in appropriate combination. Hereinafter, the configuration of the second embodiment described above will be referred to as a second basic example.

[0241] 10, three notification locations N1 and three designated locations D1 are provided. However, the number of notification locations N1 may be one, two, or four or more. The number of designated locations D1 may be one, two, or four or more.

[0242] 10, two inspection locations I1 are provided. However, the number of inspection locations I1 may be one or three or more. Furthermore, no inspection location I1 may be provided.

[0243] (Second Modification) In the example shown in Fig. 9, the inspection location I1, the designated location D1, and the notification location N1 overlap. Also, in the example shown in Fig. 10, each of the two designated locations D1, excluding the dummy designated location D1X, overlaps with the inspection location I1 and the notification location N1. Also, in the example shown in Fig. 10, each of the three designated locations D1 overlaps with the notification location N1.

[0244] However, one of the inspection location I1, the designated location D1, and the notification location N1 may be a different location from the other two. Also, the inspection location I1, the designated location D1, and the notification location N1 may be different locations from each other.

[0245] For example, if the notice 8Q is a map or sign (such as an arrow) indicating the location where the test will be conducted, the notified location N1 may be different from the designated location D1.

[0246] Also, for example, if a notice 8Q notifies that an inspection will be conducted at a location near the notice 8Q, but the notice 8Q is installed away from the inspection device 5Q, the notification location N1 is different from the inspection location I1.

[0247] Furthermore, as explained in Basic Example 2, the inspection notification may indicate a dummy designated location D1X. In other words, the inspection notification may be a notification that differs from the actual inspection situation. For example, the designated location D1 (dummy designated location D1X) notified by the notice 8Q as the location where the inspection will be conducted may be different from the actual inspection location I1. Alternatively, the notice 8Q may notify that the inspection will be conducted at the designated location D1 (dummy designated location D1X), but the inspection may not actually be conducted at any location.

[0248] (Third Modification) In this modification, an overlap between one or more designated locations D1 and one or more notification locations N1 will be described.

[0249] 10, each of the three designated locations D1 overlaps with the notification location N1. In other words, the three designated locations D1 include the three notification locations N1.

[0250] However, only some of the designated locations D1 may overlap with the notification locations N1. Therefore, the designated locations D1 may include one or more notification locations N1.

[0251] Furthermore, even when the number of designated locations D1 is one, the designated location D1 may overlap with the notification location N1.

[0252] Therefore, one or more designated locations D1 may include one or more notification locations N1.

[0253] (Fourth Modification) In this modification, an overlap between one or more designated locations D1 and one or more inspection locations I1 will be described.

[0254] 10, two of the three designated locations D1 overlap with the inspection location I1. In other words, the three designated locations D1 include two inspection locations I1.

[0255] In this way, only some of the designated locations D1 may overlap with the inspection location I1. Therefore, the designated locations D1 may include one or more inspection locations I1.

[0256] Alternatively, each of the multiple designated locations D1 may overlap with the inspection location I1.

[0257] Furthermore, even when the number of designated locations D1 is one, the designated location D1 may overlap with the inspection location I1.

[0258] Thus, one or more designated locations D1 may include one or more inspection locations I1.

[0259] (Fifth Modification) In this modification, an overlap between one or more notification locations N1 and one or more inspection locations I1 will be described.

[0260] 10, two of the three notification locations N1 overlap with the inspection location I1. In other words, the three notification locations N1 include two inspection locations I1.

[0261] In this way, only some of the notification locations N1 may overlap with the inspection locations I1. Therefore, the notification locations N1 may include one or more inspection locations I1.

[0262] Alternatively, each of the multiple notification locations N1 may overlap with the inspection location I1.

[0263] Furthermore, even when the number of notification locations N1 is one, the notification location N1 may overlap with the inspection location I1.

[0264] Therefore, one or more notification locations N1 may include one or more inspection locations I1.

[0265] (Sixth Modification) The inspection notification is not limited to being made by a notice 8Q. For example, the inspection notification may be made by a speaker emitting a sound. Specifically, the speaker may notify that an inspection to detect a violating substance is being conducted in a location around the speaker (designated location D1) by emitting a sound such as, for example, "Explosives detection is being conducted at this location." The notice 8Q and the speaker are examples of a notification unit that issues an inspection notification.

[0266] (Seventh Modification) In Basic Example 2, multiple notification locations N1 correspond one-to-one to multiple designated locations D1, and the examination notification sent at each notification location N1 indicates that the examination will be performed at the corresponding designated location D1. However, the examination notification sent at one notification location N1 may indicate that the examination will be performed at two or more designated locations D1.

[0267] (Eighth Modification) In Basic Example 2, the learning device 3Q generates a trained model that outputs a determination result based on a movement trajectory by unsupervised learning. In contrast, the learning device 3Q may generate a trained model that outputs a determination result based on a movement trajectory by supervised learning.

[0268] (Ninth Modification) The learning device 3Q may update the trained model by performing additional learning.

[0269] (Tenth Modification) The estimation unit 22Q may obtain a determination result based on the relationship between the position information of one or more designated locations D1 and the movement trajectory of a person, for example, based on a predetermined calculation formula created by a designer, without using a trained model.

[0270] (Eleventh Modification) The estimation unit 22Q may calculate the probability that the person being tested possesses or uses a violating substance, and if the probability is equal to or greater than a threshold, may output a determination result that the person being tested possesses or uses a violating substance.

[0271] (Twelfth Variation) In addition to displaying text indicating the inspection notification, the inspection notification may be provided by displaying a symbol indicating the inspection to attract attention. The inspection notification may also be provided by video, and the video may have visual effects such as color changes within the image, blinking images, moving text or symbols, etc., to attract attention.

[0272] (Thirteenth Variation) The violating substance to be inspected is not limited to a volatile substance. If the violating substance to be inspected is a non-volatile substance, the inspection device 5Q may detect the violating substance, for example, by spectroscopic inspection. More specifically, the inspection device 5Q is used together with a blower. The blower generates airflow at the inspection location I1 to send particles of the violating substance into the inspection device 5Q. The inspection device 5Q takes in the particles of the violating substance and detects the violating substance by spectroscopic inspection.

[0273] (Fourteenth Modification) An abnormality detection unit that detects an abnormal situation that disrupts a person's normal behavior may be provided in the facility. If the abnormality detection unit detects an abnormal situation, the estimation unit 22Q may stop making a determination based on the movement trajectory. In this case, the estimation unit 22Q may determine the possibility that a person possesses or uses a violating substance based on, for example, at least one of the detection information generated by the testing device 5Q and the biometric information generated by the biometric sensor 6Q.

[0274] The abnormality detection unit includes, for example, a microphone that detects sound, and detects a state in which there is someone making a loud noise in the facility as an abnormal situation. In such a case, it is highly likely that people using the facility will behave differently from usual, such as by keeping their distance from the loud noise person, and this is thought to reduce the accuracy of the determination by the estimation unit 22Q based on the movement trajectories of people.

[0275] The abnormality detection unit may include, for example, a droplet sensor for detecting rain leaks, and detect the occurrence of rain leaks as an abnormal situation. In such a case, people using the facility are likely to behave differently from normal situations, such as by moving away from the location of the rain leak, which may reduce the accuracy of the determination by the estimation unit 22Q based on the movement trajectories of people.

[0276] In addition, the estimation unit 22Q may stop making judgments based on the movement trajectory only near the location detected by the abnormality detection unit as the location where the abnormality occurred, and continue making judgments based on the movement trajectory at locations that are more than a predetermined distance away from the location where the abnormality occurred.

[0277] (Fifteenth Modification) In Basic Example 2, as described above, the computational model (trained model) used in the estimation unit 22Q is generated by machine learning. The estimation unit 22Q may be implemented as any type of artificial intelligence or system. Here, the machine learning algorithm is, for example, a neural network. However, the machine learning algorithm is not limited to a neural network and may be, for example, eXtreme Gradient Boosting (XGB) regression, Random Forest, decision tree, logistic regression, support vector machine (SVM), Naive Bayes classifier, k-nearest neighbors, or the like. Furthermore, the machine learning algorithm may be, for example, a Gaussian Mixture Model (GMM), k-means clustering, or the like.

[0278] (Sixteenth Modification) The execution entity of the test support system 1Q or test support method of the present disclosure includes a computer system. The computer system is primarily composed of a processor and memory as hardware. At least a portion of the functions of the execution entity of the test support system 1Q or test support method of the present disclosure are realized by the processor executing a program recorded in the memory of the computer system. The program may be pre-recorded in the memory of the computer system, provided via a telecommunications line, or provided by being recorded on a non-transitory recording medium readable by the computer system, such as a memory card, optical disk, or hard disk drive. The processor of the computer system is composed of one or more electronic circuits including a semiconductor integrated circuit (IC) or a large-scale integrated circuit (LSI). The integrated circuits, such as ICs and LSIs, are referred to by different names depending on the degree of integration, and include integrated circuits called system LSIs, very large-scale integration (VLSI), or ultra-large-scale integration (ULSI). Furthermore, a field-programmable gate array (FPGA), which is programmed after the LSI is manufactured, or a logic device capable of reconfiguring the connections within the LSI or the circuit partitions within the LSI, can also be employed as a processor. Multiple electronic circuits may be integrated into a single chip or distributed across multiple chips. Multiple chips may be integrated into a single device or distributed across multiple devices. The computer system referred to here includes a microcontroller having one or more processors and one or more memories. Therefore, the microcontroller is also composed of one or more electronic circuits, including a semiconductor integrated circuit or a large-scale integrated circuit.

[0279] Furthermore, it is not essential for the test support system 1Q that multiple functions of the test support system 1Q are concentrated in one housing, and multiple components of the test support system 1Q may be distributed across multiple housings. Furthermore, at least some of the functions of the test support system 1Q may be realized by a server, a cloud (cloud computing), or the like.

[0280] Conversely, in the embodiment, multiple functions distributed across multiple housings may be integrated into one housing. For example, the inspection support system 1Q may include a learning device 3Q.

[0281] (Seventeenth Modification) In Basic Example 2, the estimation unit 22Q executes a first estimation process. The first estimation process is a process of estimating the possibility that a person possesses or uses a violating substance based on the relationship between the position information of one or more designated locations D1 and the movement trajectory of the person.

[0282] In Basic Example 1, the estimation unit 23 executes a second estimation process. The second estimation process is a process for estimating the possibility that a person possesses or uses a violating substance based on a change in the person's condition before and after the person recognizes the test notification.

[0283] The estimation unit 22Q of this modification executes both the first estimation process and the second estimation process, and changes parameters related to the second estimation process based on the estimation result of the first estimation process, thereby enabling more accurate estimation by reflecting the estimation result of the first estimation process in the second estimation process.

[0284] The parameters related to the second estimation process are, for example, parameters for calculating the estimation probability from the degree of fluctuation in the state (behavior, biological information, etc.) of the person H1. The estimation unit 22Q sets the parameters so that the estimation probability increases as the possibility that the person possesses or uses a violating substance estimated in the first estimation process increases.

[0285] (18th Modification) The notice 8Q of Basic Example 2 is, for example, a display device such as a display or a projector device, and indicates that a test to detect a violating substance is being conducted. The display device may notify a person that detection of a violating substance is being conducted.

[0286] The content displayed on the display device may change over time. The processing unit 2Q of the inspection support system 1 may change the content displayed on the display device based on the estimation result of the estimating unit 22Q. For example, the display device may display at least one of the estimation result of the estimating unit 22Q and the inspection result of the inspection device 5Q, and change the display according to the progress of the estimation or inspection.

[0287] As in the first embodiment, the inspection notification by the display device may include notification of the body part of the person where the violation substance is to be detected. The inspection notification may also include notification of the body part of the person where the violation substance is currently being detected. The inspection notification may also include changing the notification of the body part of the person where the violation substance is currently being detected depending on the progress of the detection work. Furthermore, the inspection notification may also include notification that the body part of the person where the violation substance is to be detected includes internal organs.

[0288] As in the first embodiment, a restriction unit may be provided that restricts people other than the person who has been detected for a violating substance from viewing the display of the display device. For example, the restriction unit may restrict viewing of the screen of the display device from directions other than the front.

[0289] (Nineteenth Modification) In a location where the inspection device 5Q and the dummy unit 71Q are installed, the inspection notification may include at least one of blowing air onto the part of the person where the violation substance is to be detected and sucking air around the part. The inspection notification may also include shining light onto the part of the person where the violation substance is to be detected.

[0290] (Twentieth Modification) As in the first embodiment, the test support system 1Q may further include an odor providing unit that provides the odor of the violating substance around the person. The odor providing unit may provide multiple types of odors in sequence around the person.

[0291] (Summary) The above-described embodiments and the like disclose the following aspects.

[0292] An inspection support system (1; 1A; 1B; 1C; 1Q) according to a first aspect includes an information acquisition unit (21) and an estimation unit (23; 22Q). The information acquisition unit (21) acquires person information relating to the condition of a person (H1). The estimation unit (23; 22Q) estimates the possibility that the person (H1) possesses or uses a substance based on a change in the condition of the person (H1) in response to an inspection notification indicating that an inspection to detect a substance related to a crime will be conducted.

[0293] The above-described inspection support system (1; 1A; 1B; 1C; 1Q) can accurately estimate whether a person (H1) possesses or uses a substance related to a crime.

[0294] In the test support system (1; 1A) according to the second aspect, in the first aspect, the estimation unit (23) estimates the possibility that a person (H1) possesses or uses a substance based on changes in the state of the person (H1) before and after the person (H1) recognizes the test notification.

[0295] The above-mentioned testing support system (1; 1A) can accurately estimate whether a person (H1) possesses or uses a violating substance by estimating an estimated probability based on the change in the state of the person (H1) before and after the person (H1) recognizes the testing notification.

[0296] In the test support system (1) according to the third aspect, in the second aspect, the person information includes a video of the behavior of the person (H1) as the state of the person (H1). An estimation unit (23) estimates the possibility that the person (H1) possesses or uses a substance based on fluctuations in the behavior.

[0297] The above-mentioned test support system (1) can accurately estimate whether a person (H1) possesses or uses a violating substance by estimating the probability based on changes in the person's (H1) behavior before and after the person (H1) recognizes the test notification.

[0298] In a test support system (1A) according to a fourth aspect, in the second or third aspect, the person information includes biometric information of a person (H1). An estimation unit (23) estimates the possibility that the person (H1) possesses or uses a substance based on a variation in the biometric information.

[0299] The above-mentioned test support system (1A) can accurately estimate whether a person (H1) possesses or uses a violating substance by estimating an estimated probability based on fluctuations in biometric information before and after the person (H1) recognizes the test notification.

[0300] A test support system (1; 1A) according to a fifth aspect is any one of the second to fourth aspects, further comprising a timing detection unit (22). The timing detection unit (22) detects a recognition timing, which is the timing at which the person (H1) recognizes the test notification, based on a video of the person (H1). The estimation unit (23) estimates the possibility that the person (H1) is in possession of or using a substance, based on a change in the state of the person (H1) before and after the recognition timing.

[0301] The above-mentioned examination support system (1; 1A) can easily detect the recognition timing.

[0302] In the inspection support system (1; 1A) according to the sixth aspect, in the fifth aspect, the timing detection unit (22) detects the recognition timing based on the line of sight of a person (H1) appearing in the video.

[0303] The above-described inspection support system (1; 1A) can detect the recognition timing with high accuracy.

[0304] In a testing support system (1Q) according to a seventh aspect, in any one of the first to sixth aspects, the person information includes behavioral information indicating a movement trajectory of the person (H1). The testing notification indicates that a test to detect a substance will be conducted on the person (H1) at one or more designated locations (D1). The estimation unit (22Q) estimates the possibility that the person (H1) possesses or uses the substance based on the relationship between the location information of the one or more designated locations (D1) and the movement trajectory.

[0305] The above-mentioned testing support system (1Q) can accurately estimate whether a person (H1) possesses or uses a violating substance by estimating the probability based on the movement trajectory of the person (H1) when he or she recognized the testing notification.

[0306] In the examination support system (1Q) according to the eighth aspect, in the seventh aspect, the one or more designated locations (D1) include one or more notification locations (N1) where an examination notification is made.

[0307] According to the above configuration, when a person (H1) who possesses or uses a substance related to a crime receives a notification of an inspection, the person (H1) is more likely to behave suspiciously, and therefore the possibility that the person (H1) possesses or uses a substance related to a crime can be more accurately estimated.

[0308] In the examination support system (1Q) according to the ninth aspect, in the eighth aspect, the one or more designated locations (D1) are a plurality of designated locations (D1). The plurality of designated locations (D1) include a plurality of notification locations (N1) to which an examination notification is sent.

[0309] According to the above configuration, it is possible to accurately estimate the possibility that a person (H1) possesses or uses a substance related to a crime at each of a plurality of notification locations (N1). That is, it is possible to accurately estimate the possibility that a person (H1) possesses or uses a substance related to a crime over a wide range.

[0310] In the examination support system (1Q) according to the tenth aspect, in any one of the seventh to ninth aspects, the behavior information includes an image of the person (H1).

[0311] According to the above configuration, the movement trajectory of the person (H1) can be detected relatively easily.

[0312] An inspection support system (1Q) according to an eleventh aspect is the inspection support system (1Q) of any one of the seventh to tenth aspects, further comprising a detection acquisition unit (23Q). The detection acquisition unit (23Q) acquires detection information representing detection results of crime-related substances at one or more inspection locations (I1). The estimation unit (22Q) estimates the possibility that a person (H1) possesses or uses a crime-related substance based on the detection information.

[0313] According to the above configuration, it is possible to more accurately estimate the possibility that a person (H1) possesses or uses a substance related to a crime.

[0314] In the inspection support system (1Q) according to the twelfth aspect, in the eleventh aspect, the detection information is information representing the detection result of a volatile substance as a substance related to a crime.

[0315] According to the above configuration, it is possible to accurately estimate the possibility that a person (H1) possesses or uses a volatile substance (explosive, specific drug, etc.) related to a crime.

[0316] In the inspection support system (1Q) according to the thirteenth aspect, in the eleventh or twelfth aspect, the detection information includes information representing the concentration of a substance involved in crime in the air.

[0317] According to the above configuration, it is possible to accurately estimate the possibility that a person (H1) possesses or uses a volatile substance (explosive, specific drug, etc.) related to a crime.

[0318] In the examination support system (1Q) according to the fourteenth aspect, in any one of the eleventh to thirteenth aspects, the one or more designated locations (D1) include one or more examination locations (I1).

[0319] According to the above configuration, the possibility that a person (H1) possesses or uses a substance related to a crime can be estimated more accurately compared to when one or more inspection locations (I1) are different from one or more designated locations (D1).

[0320] A test support system (1Q) according to a fifteenth aspect is any one of the seventh to fourteenth aspects, further including a biometric information acquisition unit (24Q). The biometric information acquisition unit (24Q) acquires biometric information of a person (H1). The estimation unit (22Q) estimates the possibility that the person (H1) possesses or uses a substance related to a crime, based on the biometric information.

[0321] According to the above configuration, it is possible to more accurately estimate the possibility that a person (H1) possesses or uses a substance related to a crime.

[0322] In the inspection support system (1Q) according to the 16th aspect, in the 15th aspect, the estimation unit (22Q) estimates the possibility that a person (H1) possesses or uses a substance related to a crime, further based on the difference between biometric information when the person (H1) is present at one or more notification locations (N1) where an inspection notification is issued and biometric information when the person (H1) is present at a location other than the one or more notification locations (N1).

[0323] According to the above configuration, it is possible to detect changes in the biometric information of the person (H1) due to, for example, the person being upset after receiving the test notification, and therefore it is possible to more accurately estimate the possibility that the person (H1) possesses or uses a substance related to a crime.

[0324] In the inspection support system (1Q) relating to the 17th aspect, in the 15th or 16th aspect, the estimation unit (22Q) estimates the possibility that a person (H1) possesses or uses a substance related to a crime, further based on the difference between biometric information when the person (H1) is present in one or more designated locations (D1) and biometric information when the person (H1) is present in a location other than the one or more designated locations (D1).

[0325] According to the above configuration, it is possible to detect changes in the biometric information of the person (H1) due to, for example, shaking while passing through the designated location (D1), and therefore it is possible to more accurately estimate the possibility that the person (H1) possesses or uses a substance related to a crime.

[0326] In an inspection support system (1Q) according to an eighteenth aspect, in any one of the seventh to seventeenth aspects, the behavior information includes at least one of direction information indicating a direction of movement of the person (H1) and distance information indicating a distance between the person (H1) and one or more designated locations (D1). The estimation unit (22Q) estimates the possibility that the person (H1) possesses or uses a substance related to a crime, based on at least one of the direction information and the distance information.

[0327] According to the above configuration, it is possible to more accurately estimate the possibility that a person (H1) possesses or uses a substance related to a crime.

[0328] In the inspection support system (1Q) relating to the 19th aspect, in any one of the 7th to 18th aspects, one or more designated locations (D1) include a dummy designated location (D1X) where inspections to detect substances related to crimes are not actually conducted.

[0329] According to the above configuration, the number of pieces of equipment (inspection device (5Q)) required for inspection to detect crime-related substances can be reduced.

[0330] In a test support system (1; 1A; 1B; 1C; 1Q) according to a twentieth aspect, in any one of the seventh to nineteenth aspects, the estimation unit (23:22Q) executes a first estimation process and a second estimation process. In the first estimation process, the possibility that a person (H1) possesses or uses a substance is estimated based on the relationship between the position information of one or more designated locations (D1) and the movement trajectory. In the second estimation process, the possibility that a person (H1) possesses or uses a substance is estimated based on changes in the person (H1)'s state before and after the person (H1) recognizes the test notification. The estimation unit (23:22Q) changes parameters related to the second estimation process based on the estimation result of the first estimation process.

[0331] According to the above configuration, it is possible to more accurately estimate the possibility that a person (H1) possesses or uses a substance related to a crime.

[0332] The test support system (1; 1C; 1Q) according to the 21st aspect is any one of the first to 20th aspects, and further includes a sensor (such as an odor sensor 8) that detects substances by analyzing volatile components.

[0333] According to the above configuration, an inspection support system (1; 1A; 1B; 1C; 1Q) equipped with a sensor can be provided.

[0334] A test support system (1; 1C; 1Q) according to a 22nd aspect is any one of the first to 21st aspects, further comprising a setting unit (25; 26Q). The setting unit (25; 26Q) changes the setting of a sensor (such as an odor sensor 8) that detects a substance by analyzing volatile components, based on the estimation result of the estimation unit (23; 22Q).

[0335] According to the above configuration, the detection accuracy of the sensor can be improved by changing the settings of the sensor.

[0336] The test support system (1; 1A; 1B; 1C; 1Q) according to the 23rd aspect is any one of the first to 22nd aspects, and further includes a memory unit (11; 12Q) that stores person information.

[0337] According to the above configuration, personal information can be confirmed after the fact.

[0338] The inspection support system (1C; 1Q) according to the 24th aspect, in any one of the first to 23rd aspects, further comprises an intake port (opening 1220; opening 520Q; 70Q) that indicates to a person (H1) that substance detection is being performed.

[0339] According to the above configuration, the violator is made aware that a substance is being inhaled, and this can induce agitation in the violator.

[0340] In the inspection support system (1; 1A; 1B; 1C; 1Q) according to the 25th aspect, in any one of the first to 24th aspects, the inspection notification includes displaying on the display device (10) that substance detection is being performed on the person (H1).

[0341] According to the above configuration, the person (H1) can be made aware that substance detection is currently being performed.

[0342] The test support system (1B; 1Q) according to the 26th aspect is the 25th aspect, and further includes a restriction unit (112) that restricts people other than the person (H1) undergoing substance detection from viewing the display on the display device (10).

[0343] According to the above configuration, it is possible to make the person (H1) have a unique reaction when the person (H1) sees the display of the display device (10) for the first time.

[0344] The examination support system (1; 1A; 1B; 1C; 1Q) according to the 27th aspect, in the 25th or 26th aspect, further includes a display control unit (26) that changes the content displayed on the display device (10) based on the estimation result of the estimation unit (23:22Q).

[0345] According to the above configuration, it is possible to induce agitation in the violator.

[0346] In the test support system (1B; 1Q) according to the 28th aspect, in any one of the first to 27th aspects, the test notification includes notification of the part of the person (H1) where the substance is to be detected.

[0347] According to the above configuration, when the violator is notified that a substance is being detected in the area where the violating substance is being hidden, the violator is likely to become agitated and change his / her behavior, thereby improving the estimation accuracy of the estimation unit (23:22Q).

[0348] In the inspection support system (1B; 1Q) according to the twenty-ninth aspect, in the twenty-eighth aspect, the inspection notification includes notification of the part of the person (H1) where the substance is currently being detected.

[0349] According to the above configuration, when a violator is notified that a substance is being detected in a part of the body where the violating substance is being hidden, the violator is likely to become agitated and change his / her behavior, thereby improving the estimation accuracy of the estimation unit (23:22Q).

[0350] In the inspection support system (1B; 1Q) according to the 30th aspect, in the 29th aspect, the inspection notification includes changing the notification of the part of the person (H1) where the substance is currently being detected according to the progress of the detection work.

[0351] According to the above configuration, when a violator is notified that a substance is being detected in a part of the body where the violating substance is being hidden, the violator is likely to become agitated and change his / her behavior, thereby improving the estimation accuracy of the estimation unit (23:22Q).

[0352] In the test support system (1B; 1Q) according to the 31st aspect, in any one of the 28th to 30th aspects, the test notification includes a notification that the part of the person (H1) where the substance is to be detected includes internal organs.

[0353] According to the above configuration, it is possible to improve the accuracy of estimation when the violator is hiding a violating substance in his internal organs.

[0354] In the inspection support system (1C; 1Q) according to the 32nd aspect, in any one of the 28th to 31st aspects, the inspection notification includes at least one of blowing air onto the target part of the person (H1) where the substance is to be detected, and sucking in the air around the target part.

[0355] According to the above configuration, when the violator is exposed to wind on the area where the violating substance is hidden or when the air around the area where the violating substance is hidden is sucked in, the violator is likely to become agitated and change his / her behavior, thereby improving the estimation accuracy of the estimation unit (23; 22Q).

[0356] In the test support system (1B; 1Q) according to the 33rd aspect, in any one of the 28th to 32nd aspects, the test notification includes shining light onto a part of the person (H1) where the substance is to be detected.

[0357] According to the above configuration, when a light is shone on the part of the body where the violator is hiding the violating substance, the violator is likely to become agitated and change his / her behavior, thereby improving the estimation accuracy of the estimation unit (23; 22Q).

[0358] The test support system (1C; 1Q) according to the 34th aspect is any one of the first to 33rd aspects, and further includes an odor providing unit (13) that provides the odor of a substance around the person (H1).

[0359] According to the above configuration, it is possible to induce agitation in the violator.

[0360] In the test support system (1C; 1Q) according to the 35th aspect, in the 34th aspect, the odor providing unit (13) sequentially provides multiple types of odors, including the odor of a substance, around a person (H1).

[0361] According to the above configuration, it is possible to induce agitation in the violator.

[0362] In an inspection support system (1; 1A; 1B; 1C; 1Q) according to a 36th aspect, in any one of the first to 35th aspects, an estimation unit (23; 22Q) estimates the possibility that a person (H1) possesses or uses a substance related to a crime, using a trained model generated by machine learning.

[0363] According to the above configuration, it is possible to more accurately estimate the possibility that a person (H1) possesses or uses a substance related to a crime.

[0364] In a testing support system (1) according to a thirty-seventh aspect, in the thirty-sixth aspect, the person information includes behavioral information indicating a movement trajectory of a person (H1). The testing notification indicates that a test to detect a substance will be conducted on the person (H1) at one or more specified locations (D1). The trained model outputs a result indicating that the greater the degree to which the movement trajectory of the person (H1) to be tested deviates from the average movement trajectory of multiple people (H1), the more likely the person (H1) to be tested possesses or uses a substance related to a crime.

[0365] According to the above configuration, compared to performing machine learning using training data (teacher data) that contains correct answer data on whether or not a person actually possessed a substance related to a crime, it is easier to collect training data, making it easier to generate a trained model.

[0366] The examination support system (1; 1A; 1B; 1C; 1Q) according to the thirty-eighth aspect is any one of the first to thirty-seventh aspects, and further comprises a human information generation unit (3) that generates human information.

[0367] The above-described inspection support system (1; 1A; 1B; 1C; 1Q) can accurately estimate whether a person (H1) possesses or uses a substance related to a crime.

[0368] In the examination support system (1; 1A; 1B; 1C; 1Q) according to the thirty-ninth aspect, in any one of the first to thirty-eighth aspects, the examination notification is visually displayed to the person (H1).

[0369] The above-described inspection support system (1; 1A; 1B; 1C; 1Q) can accurately estimate whether a person (H1) possesses or uses a substance related to a crime.

[0370] The examination support system (1; 1A; 1B; 1C; 1Q) according to the fortieth aspect is the thirty-ninth aspect, further comprising a display unit (4) that displays an examination notification.

[0371] The above-described inspection support system (1; 1A; 1B; 1C; 1Q) can accurately estimate whether a person (H1) possesses or uses a substance related to a crime.

[0372] The inspection support system (1; 1A; 1B; 1C; 1Q) according to the 41st aspect is any one of the 1st to 40th aspects, and further includes a result output unit (24) that outputs the estimation result of the estimation unit (23; 22Q).

[0373] The above-described examination support system (1; 1A; 1B; 1C; 1Q) can output the estimation result of the estimation unit (23; 22Q) to the outside.

[0374] In the inspection support system (1; 1A; 1B; 1C; 1Q) according to the forty-second aspect, in any one of the first to forty-first aspects, the substance includes at least one of a drug and an explosive.

[0375] The above-described inspection support system (1; 1A; 1B; 1C; 1Q) can accurately estimate whether a person (H1) possesses or uses at least one of drugs and explosives.

[0376] In the test support system (1; 1A; 1B; 1C; 1Q) relating to the 43rd aspect, in any one of the first to 42nd aspects, the test notification indicates that a test will be conducted on a person (H1) to detect a substance by analyzing volatile components.

[0377] The above-described inspection support system (1; 1A; 1B; 1C; 1Q) can accurately estimate whether a person (H1) possesses or uses at least one of drugs and explosives.

[0378] The configurations other than the first aspect are not essential for the examination support system (1; 1A; 1B; 1C; 1Q) and can be omitted as appropriate.

[0379] The inspection assistance method according to the forty-fourth aspect includes an information acquisition step (S1) and an estimation step (S3). In the information acquisition step (S1), personal information relating to the condition of a person (H1) is acquired. In the estimation step (S3), a possibility that the person (H1) possesses or uses a substance is estimated based on a change in the condition of the person (H1) in response to an inspection notification indicating that an inspection to detect a substance related to a crime will be conducted.

[0380] The above-described inspection assistance method can accurately estimate whether or not a person (H1) possesses or uses a substance related to a crime.

[0381] A program according to a forty-fifth aspect causes a computer system to execute the examination support method of the forty-fourth aspect.

[0382] The above program can accurately estimate whether a person (H1) possesses or uses a substance related to a crime.

[0383] Not limited to the above aspects, various configurations (including modified examples) of the inspection support system (1; 1A; 1B; 1C; 1Q) according to the embodiment can be embodied as an inspection support method, a (computer) program, or a non-transitory recording medium on which a program is recorded.

[0384] 1; 1A; 1B; 1C; 1Q Test support system 10 Display device 11 Memory unit 13 Scent providing unit 112 Restriction unit 21 Information acquisition unit 22 Timing detection unit 23 Estimation unit 24 Result output unit 25 Setting unit 26 Display control unit 3 Person information generation unit 4 Display unit H1 Person S1 Information acquisition step S3 Estimation step S4 Result output step 12Q Memory unit 22Q Estimation unit 23Q Detection acquisition unit 24Q Biometric information acquisition unit 26Q Setting unit D1 Designated location D1X Dummy designated location I1 Test location N1 Notification location

Claims

1. An inspection support system comprising: an information acquisition unit that acquires human information relating to a person's condition; and an estimation unit that estimates the possibility that the person possesses or uses a substance based on a change in the person's condition in response to an inspection notification indicating that an inspection to detect a substance related to a crime will be conducted.

2. The testing support system of claim 1, wherein the estimation unit estimates the possibility that the person possesses or uses the substance based on a change in the person's condition before and after the person recognizes the testing notification.

3. The inspection support system of claim 2, wherein the human information includes a video of the person's behavior as the person's condition, and the estimation unit estimates the possibility that the person is in possession of or using the substance based on changes in the behavior.

4. The testing support system of claim 2 or 3, wherein the personal information includes biometric information of the person, and the estimation unit estimates the possibility that the person possesses or uses the substance based on fluctuations in the biometric information.

5. A testing support system according to any one of claims 2 to 4, further comprising a timing detection unit that detects a recognition timing, which is the timing at which the person recognizes the testing notification, based on a video of the person, and the estimation unit estimates the possibility that the person possesses or uses the substance based on fluctuations in the person's condition before and after the recognition timing.

6. The inspection support system according to claim 5, wherein the timing detection unit detects the recognition timing based on the line of sight of the person appearing in the video.

7. An inspection support system according to any one of claims 1 to 6, wherein the personal information includes behavioral information indicating the person's movement trajectory, the inspection notification indicates that an inspection to detect the substance will be conducted on the person at one or more designated locations, and the estimation unit estimates the possibility that the person possesses or uses the substance based on the relationship between the location information of the one or more designated locations and the movement trajectory.

8. The inspection support system according to claim 7, wherein the one or more designated locations include one or more notification locations to which the inspection notification is to be sent.

9. The inspection support system according to claim 8, wherein the one or more designated locations are a plurality of designated locations, and the plurality of designated locations include a plurality of notification locations to which the inspection notification is to be sent.

10. The examination support system according to any one of claims 7 to 9, wherein the behavioral information includes an image of the person.

11. An inspection support system according to any one of claims 7 to 10, further comprising a detection acquisition unit that acquires detection information representing the detection results of the substance related to a crime at one or more inspection locations, and the estimation unit further estimates the possibility that the person possesses or uses the substance based on the detection information.

12. The inspection support system according to claim 11, wherein the detection information is information that represents the detection result of a volatile substance as the substance related to a crime.

13. The inspection support system according to claim 11 or 12, wherein the detection information includes information representing a concentration of the substance in the air that is related to a crime.

14. The inspection support system according to any one of claims 11 to 13, wherein the one or more designated locations include the one or more inspection locations.

15. An examination support system according to any one of claims 7 to 14, further comprising a biometric information acquisition unit that acquires biometric information of the person, and the estimation unit further estimates the possibility that the person possesses or uses the substance based on the biometric information.

16. The testing support system of claim 15, wherein the estimation unit estimates the possibility that the person possesses or uses the substance based further on a difference between the biometric information when the person is present at one or more notification locations where the testing notification is sent and the biometric information when the person is present at a location other than the one or more notification locations.

17. The testing support system of claim 15 or 16, wherein the estimation unit estimates the possibility that the person possesses or uses the substance based further on a difference between the biometric information when the person is present at the one or more designated locations and the biometric information when the person is present at a location other than the one or more designated locations.

18. An inspection support system according to any one of claims 7 to 17, wherein the behavioral information includes at least one of directional information representing the direction of movement of the person and distance information representing the distance between the person and the one or more designated locations, and the estimation unit estimates the possibility that the person possesses or uses the substance based on at least one of the directional information and the distance information.

19. The inspection support system of any one of claims 7 to 18, wherein the one or more designated locations include a dummy designated location where the inspection to detect the substance related to a crime is not actually carried out.

20. The testing support system of any one of claims 7 to 19, wherein the estimation unit executes a first estimation process to estimate the possibility that the person possesses or uses the substance based on the relationship between the location information of the one or more designated locations and the movement trajectory, and a second estimation process to estimate the possibility that the person possesses or uses the substance based on changes in the person's condition before and after the person recognizes the testing notification, and the estimation unit changes parameters related to the second estimation process based on the estimation result of the first estimation process.

21. The inspection support system according to any one of claims 1 to 20, further comprising a sensor that detects the substance by analyzing a volatile component.

22. The inspection support system according to any one of claims 1 to 21, further comprising a setting unit that changes the settings of a sensor that detects the substance by analyzing volatile components based on the estimation result of the estimation unit.

23. The inspection support system according to any one of claims 1 to 22, further comprising a storage unit that stores the personal information.

24. The inspection support system of any one of claims 1 to 23, further comprising an intake port that indicates to the person that detection of the substance is occurring.

25. The inspection support system according to any one of claims 1 to 24, wherein the inspection notification includes displaying on a display device to the person that detection of the substance is being performed.

26. The inspection support system according to claim 25, further comprising a restriction unit that restricts people other than the person undergoing detection of the substance from viewing the display on the display device.

27. The inspection support system according to claim 25 or 26, further comprising a display control unit that changes the content displayed on the display device based on the estimation result of the estimation unit.

28. The examination support system of any one of claims 1 to 27, wherein the examination notification includes notification of a part of the person where the substance is to be detected.

29. The examination support system of claim 28, wherein the examination notification includes notification of the part of the person where the substance is currently being detected.

30. The inspection support system according to claim 29, wherein the inspection notification includes changing the notification of the part of the person where the substance is currently being detected according to the progress of the detection work.

31. The examination support system according to any one of claims 28 to 30, wherein the examination notification includes a notification that the part of the person where the substance is to be detected includes an internal organ.

32. An inspection support system according to any one of claims 28 to 31, wherein the inspection notification includes at least one of blowing air onto the target part of the person where the substance is to be detected, and sucking in air around the target part.

33. The examination support system according to any one of claims 28 to 32, wherein the examination notification includes shining light on the target part of the person where the substance is to be detected.

34. The inspection support system according to any one of claims 1 to 33, further comprising an odor providing unit that provides the odor of the substance around the person.

35. The inspection support system according to claim 34, wherein the odor providing unit sequentially provides a plurality of types of odor including the odor of the substance to the surroundings of the person.

36. The testing support system of any one of claims 1 to 35, wherein the estimation unit estimates the possibility that the person possesses or uses the substance using a trained model generated by machine learning.

37. The testing support system of claim 36, wherein the personal information includes behavioral information indicating the movement trajectory of the person, the testing notification indicates that a test to detect the substance will be conducted on the person at one or more designated locations, and the trained model outputs a result that the greater the degree to which the movement trajectory of the person to be tested deviates from the average of the movement trajectories of multiple people, the higher the likelihood that the person to be tested possesses or uses the substance.

38. The inspection support system according to any one of claims 1 to 37, further comprising a human information generating unit that generates the human information.

39. The examination support system according to any one of claims 1 to 38, wherein the examination notification is visually displayed to the person.

40. The examination support system according to claim 39, further comprising a display unit that displays the examination notice.

41. The inspection support system according to any one of claims 1 to 40, further comprising a result output unit that outputs the estimation result of the estimation unit.

42. The inspection support system of any one of claims 1 to 41, wherein the substance includes at least one of a drug and an explosive.

43. The testing support system of any one of claims 1 to 42, wherein the testing notification indicates that a test that detects the substance by analyzing volatile components will be conducted on the person.

44. An inspection assistance method comprising: an information acquisition step of acquiring human information relating to a person's condition; and an estimation step of estimating the possibility that the person possesses or uses a substance based on a change in the person's condition in response to a testing notification indicating that a test to detect a substance related to a crime will be conducted.

45. A program for causing a computer system to execute the inspection support method of claim 44.

Images