Person detection device, person detection system, person detection method, and person detection program

Abstract

A person detection device (100) comprises: an acquisition unit (110) that acquires a captured video image; an object detection unit (120) that detects a prescribed object to be detected from the video image on the basis of a threshold value of a prescribed object detection score; an adjustment unit (130) that lowers a threshold value of a person detection score in a region (R2) of a prescribed range which includes the object to be detected in the video image to a value lower than threshold values of person detection scores in other regions; and a person detection unit (140) that detects a person from the video image on the basis of the threshold value of the person detection score.

Description

Person detection device, person detection system, person detection method, and non-transitory computer-readable medium

[0001] The present disclosure relates to a person detection device, a person detection system, a person detection method, and a non-transitory computer-readable medium.

[0002] When a person or the like is captured in video captured by a vehicle-mounted imaging device such as a drive recorder, the area in which the person is captured may be masked from the perspective of protecting personal information. In order to mask the area in which the person is captured in the video, it is necessary to accurately detect the person from the captured video.

[0003] Patent Literature 1 describes a technology for detecting people from an image captured by an imaging device. Specifically, Patent Literature 1 describes a technology for dividing a captured image into congested areas and quiet areas, acquiring a person determination threshold corresponding to the congestion level for each area to generate a threshold map, and performing person detection for each of a plurality of areas using the person determination threshold corresponding to each area based on the threshold map.

[0004] Japanese Patent Application Laid-Open No. 2017-097510

[0005] Lowering the threshold for detecting people in order to improve the accuracy of detecting people from video may result in masking areas in the video that should not be masked, which is undesirable. On the other hand, lowering the threshold for detecting people may result in a decrease in the accuracy of detecting people from video. Patent Document 1 describes changing the person determination threshold for each area according to the congestion level, but does not perform person detection for masking, and therefore cannot solve the problem.

[0006] The present disclosure has been made to solve such problems, and aims to provide a human detection device, a human detection system, a human detection method, and a non-transitory computer-readable medium that are capable of detecting people accurately and suitably.

[0007] A human detection device according to a first aspect of the present disclosure includes an acquisition means for acquiring a captured image, an object detection means for detecting a predetermined detection object from the image based on a predetermined object detection score threshold, an adjustment means for lowering the threshold of the human detection score in a predetermined range of areas in the image that includes the detection object below the threshold of the human detection score in other areas, and a human detection means for detecting a human from the image based on the threshold of the human detection score.

[0008] A person detection system according to a second aspect of the present disclosure comprises an imaging device installed in a vehicle and capturing video of the area around the vehicle, and a person detection device capable of communicating with the imaging device, wherein the person detection device comprises an acquisition means for acquiring video captured by the imaging device, an object detection means for detecting a predetermined detection target from the video based on a predetermined target detection score threshold, an adjustment means for lowering the threshold of the person detection score in a predetermined range of areas within the video that includes the detection target below the threshold of the person detection score in other areas, and a person detection means for detecting a person from the video based on the threshold of the person detection score.

[0009] A person detection method according to a third aspect of the present disclosure is a method in which a computer acquires a captured video, detects a predetermined detection target from the video based on a predetermined target detection score threshold, lowers the person detection score threshold in a predetermined range of areas within the video that includes the detection target below the thresholds of the person detection score in other areas, and detects a person from the video based on the threshold of the person detection score.

[0010] A non-transitory computer-readable medium according to a fourth aspect of the present disclosure stores a person detection program that causes a computer to perform the following processes: acquiring a captured image; detecting a predetermined detection target from the image based on a predetermined target detection score threshold; lowering the person detection score threshold in a predetermined range of areas in the image that includes the detection target below the thresholds of the person detection score in other areas; and detecting a person from the image based on the threshold of the person detection score.

[0011] It is possible to provide a human detection device, a human detection system, a human detection method, and a non-transitory computer-readable medium that are capable of detecting humans accurately and suitably.

[0012] FIG. 1 is a block diagram showing the configuration of a human detection device according to embodiment 1. FIG. 2 is a diagram showing an example of video acquired by an acquisition unit according to embodiment 1. FIG. 3 is a flowchart showing a human detection method according to embodiment 1. FIG. 4 is a block diagram showing the configuration of a human detection system according to embodiment 2. FIG. 5 is a block diagram showing the configuration of a human detection device according to embodiment 2. FIG. 6 is a diagram showing an example of video acquired by an acquisition unit according to embodiment 2. FIG. 7 is a flowchart showing a human detection method according to embodiment 2. FIG. 8 is a block diagram showing the configuration of a human detection device according to embodiment 3. FIG. 9 is a diagram showing an example of video acquired by an acquisition unit according to embodiment 3. FIG. 10 is a flowchart showing a human detection method according to embodiment 3.

[0013] Embodiments of the present disclosure will be described in detail below with reference to the drawings. In each drawing, the same or corresponding elements are denoted by the same reference numerals, and redundant description will be omitted as necessary for clarity of description. <Embodiment 1> FIG. 1 is a block diagram showing the configuration of a human detection device 100 according to embodiment 1. The human detection device 100 includes an acquisition unit 110 as an acquisition means, an object detection unit 120 as an object detection means, an adjustment unit 130 as an adjustment means, and a human detection unit 140 as a human detection means. The human detection device 100 is connected to a network 500 (not shown). The network 500 may be wired or wireless. An image capture device 310 (not shown) and the like are connected to the network 500. The image capture device 310 is installed in a vehicle 300 (not shown) and captures images of the surroundings of the vehicle 300. The video captured by the image capture device 310 is a video and includes multiple consecutive frames arranged in chronological order. Here, a frame refers to still image data captured by the image capture device 310.

[0014] The acquisition unit 110 acquires video captured by an image capturing device 310 installed in the vehicle 300. The video includes a plurality of frames. The video captured by the image capturing device 310 is transmitted from the image capturing device 310 to the human detection device 100 via the network 500. The video acquired by the acquisition unit 110 may be video captured by an image capturing device other than the image capturing device 310 installed in the vehicle 300, such as a surveillance camera.

[0015] The object detection unit 120 detects a predetermined detection target from the video acquired by the acquisition unit 110 based on a predetermined object detection score threshold. Examples of the predetermined detection target include moving objects such as motorcycles and four-wheeled vehicles. In this first embodiment, a motorcycle is used as an example of the predetermined detection target. Specifically, the object detection unit 120 detects a motorcycle for each frame included in the video based on a motorcycle detection score threshold as the predetermined object detection score. The object detection unit 120 detects a motorcycle using a trained motorcycle detection model (not shown). The motorcycle detection score is a score calculated by the object detection unit 120 using the motorcycle detection model for each detection frame area of ​​a predetermined size (e.g., M × N pixels (M and N are integers equal to or greater than 2)) from the frame. The motorcycle detection score is higher in areas where a motorcycle is likely to be present than in other areas. The predetermined motorcycle detection score threshold is a preset value and is used when detecting a motorcycle from frames constituting the video. If the motorcycle detection score for an area within a frame is equal to or greater than the motorcycle detection score threshold, the object detection unit 120 determines that a motorcycle is present in the area. Here, motorcycles include motorcycles, bicycles, and electric kick scooters. Note that, in this specification, machine learning may be deep learning, but is not limited thereto. Furthermore, the method for calculating the motorcycle detection score is not limited to the above, and other existing technologies may be applied.

[0016] When the object detection unit 120 detects a motorcycle from a frame constituting a video, the adjustment unit 130 lowers the threshold value of the human detection score in a predetermined range of an area in the frame that includes the motorcycle. Here, the predetermined range is a range of a predetermined size that is set appropriately depending on the purpose of human detection.

[0017] The person detection unit 140 detects people from the video acquired by the acquisition unit 110 based on a predetermined person detection score threshold. Specifically, the person detection unit 140 detects people for each frame included in the video. The person detection unit 140 detects people using a trained person detection model (not shown). The person detection score is a score calculated by the person detection unit 140 using the person detection model for each detection frame region of a predetermined size (e.g., M × N pixels (M and N are integers equal to or greater than 2)) from the frame. The person detection score is higher in regions where a person is likely to be present than in other regions. The predetermined person detection score threshold is a preset value and is used when detecting people from frames constituting the video. The person detection score threshold may be set to a different value depending on the region within the frame. Specifically, the person detection score threshold is adjusted by the adjustment unit 130 in predetermined cases. If the person detection score for a region within a frame is equal to or greater than the person detection score threshold, the person detection unit 140 determines that a person is present in the region. The method for calculating the human detection score is not limited to the above, and other existing techniques can be applied.

[0018] FIG. 2 shows an example of a frame 10 included in a video acquired by the acquisition unit 110. The frame 10 shown in FIG. 2 shows a person 30 riding a bicycle 20. In the example shown in FIG. 2, the object detection unit 120 detects the bicycle 20 as a two-wheeled vehicle. In FIG. 2, a region R1 corresponding to the bicycle 20 detected by the object detection unit 120 is indicated by a two-dot chain line. Next, the adjustment unit 130 lowers the threshold of the person detection score in a region R2 of a predetermined range that includes the region R1. In FIG. 2, the region R2 of the predetermined range is indicated by a dashed dot line. Next, the person detection unit 140 detects a person from the frame. Note that the threshold of the person detection score for region R2 is lowered compared to the thresholds of the person detection scores for the other regions, making it easier to detect a person in region R2 than in the other regions. In FIG. 2, a region R3 corresponding to a person 30 detected by the person detection unit 140 is indicated by a dashed line.

[0019] Next, the person detection method according to the first embodiment will be described with reference to FIG. 3 . First, the acquisition unit 110 acquires video captured by a camera (step S101). Next, the object detection unit 120 detects a motorcycle from a frame included in the video based on a predetermined motorcycle detection score threshold (step S102). Next, the adjustment unit 130 adjusts the person detection score threshold (step S103). Specifically, when a motorcycle is detected in the video, the adjustment unit 130 lowers the person detection score threshold in a predetermined region R2 that includes the motorcycle in the frame. Next, the person detection unit 140 detects a person from the frame that constitutes the video based on the predetermined person detection score threshold (step S104). Here, the person detection score threshold in the predetermined region R2 that includes the motorcycle is lower than the person detection score thresholds in other regions. Therefore, the person detection unit 140 can accurately detect the person 30 riding the motorcycle 20. This makes it possible to accurately detect the person 30 in the region R2 that includes the motorcycle 20 and perform masking processing from the perspective of protecting personal information. At the same time, it is possible to avoid the inconvenience of erroneously detecting a person and performing masking processing in a region other than the region R2 that includes the motorcycle 20.

[0020] In this way, the human detection device 100 according to this embodiment detects people by lowering the threshold of the human detection score in an area where a person is likely to be captured (the predetermined area R2 including the motorcycle), and therefore can accurately detect people. Furthermore, the threshold of the human detection score cannot be lowered in an area where a person is unlikely to be captured (an area other than the area R2 including the motorcycle 20), and therefore it is possible to avoid the inconvenience of erroneously detecting a person in that area. Therefore, the human detection device 100 according to this embodiment can accurately and suitably detect people.

[0021] The human detection device 100 includes a processor, a memory, and a storage device (not shown). The storage device stores a computer program that implements the processing of the human detection method according to this embodiment. The processor then loads the computer program from the storage device into the memory and executes the computer program. As a result, the processor realizes the functions of the acquisition unit 110, the object detection unit 120, the adjustment unit 130, and the human detection unit 140.

[0022] Furthermore, the acquisition unit 110, the object detection unit 120, the adjustment unit 130, and the person detection unit 140 may each be realized by dedicated hardware. Furthermore, some or all of the components of each device may be realized by general-purpose or dedicated circuits, processors, etc., or a combination thereof. These may be configured by a single chip, or by multiple chips connected via a bus. Some or all of the components of each device may be realized by a combination of the above-mentioned circuits, etc., and a program. Furthermore, a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), an FPGA (Field-Programmable Gate Array), etc. may be used as the processor.

[0023] Furthermore, when some or all of the components of the human detection device 100 are realized by multiple information processing devices, circuits, etc., the multiple information processing devices, circuits, etc. may be centrally or decentralized. For example, the information processing devices, circuits, etc. may be realized as a client-server system, a cloud computing system, or the like, in a form in which each is connected via a communication network. Furthermore, the functions of the human detection device 100 may be provided in a SaaS (Software as a Service) format.

[0024] 4 is a block diagram showing the configuration of a person detection system 200 according to embodiment 2. The person detection system 200 includes at least a photographing device 310 and a person detection device 400, and may further include a recording device 320. The photographing device 310 and the recording device 320 are each connected to the person detection device 400 via a network 500. Note that descriptions that overlap with embodiment 1 will be omitted as appropriate.

[0025] The person detection system 200 is a system for detecting people from video captured by a vehicle 300. The vehicle 300 is, for example, an automobile, but may also be a vehicle other than an automobile, such as a motorcycle or a bicycle. The vehicle 300 is equipped with an imaging device 310 and a recording device 320. The imaging device 310 is a device that captures the scenery around the vehicle 300, such as a drive recorder. The imaging device 310 includes an imaging unit 311 and a communication unit 312. The imaging unit 311 is a camera. The imaging unit 311 captures, for example, the scenery ahead of the vehicle 300, i.e., the scenery that can be seen by a driver seated in the driver's seat of the vehicle 300. The communication unit 312 is a communication interface with the network 500. The communication unit 312 transmits the video captured by the imaging unit 311 to the person detection device 400 via the network 500.

[0026] Recording device 320 is a device that records the traveling speed of vehicle 300. Recording device 320 includes a measurement unit 321 and a communication unit 322. Measurement unit 321 measures the traveling speed of vehicle 300. Communication unit 322 is a communication interface with network 500. Communication unit 322 transmits speed information including the speed measured by measurement unit 321 to human detection device 400 via network 500.

[0027] Next, the configuration of the human detection device 400 will be described in detail with reference to Fig. 5. Fig. 5 is a block diagram showing the configuration of the human detection device 400. The human detection device 400 includes a memory 410, a communication unit 420, a storage unit 430, and a control unit 440.

[0028] The memory 410 is a storage area that temporarily stores the processing contents of the control unit 440, and is a volatile storage device such as a RAM (Random Access Memory). The communication unit 420 is an interface that communicates with the outside of the human detection device 400. The storage unit 430 is a storage device that stores a human detection score threshold 431, a program 432, and the like. The human detection score threshold 431 is a numerical value used when detecting a person from a frame included in a video, and may be set to a different value depending on the area within the frame. Specifically, the human detection score threshold is adjusted by the adjustment unit 442 in predetermined cases. The program 432 is a computer program in which the human detection process according to this embodiment is implemented.

[0029] The control unit 440 includes an acquisition unit 441, an adjustment unit 442, a person detection unit 443, and a masking unit 444. The control unit 440 is a control device that controls the operation of the person detection device 400, and is, for example, a processor such as a CPU. The control unit 440 loads the program 432 from the storage unit 430 into the memory 410 and executes it. In this way, the control unit 440 realizes the functions of the acquisition unit 441, the adjustment unit 442, the person detection unit 443, and the masking unit 444.

[0030] The acquisition unit 441 acquires the video transmitted from the imaging device 310. The video includes multiple consecutive frames. The video may also include identification information, time information, etc. The identification information is information for identifying the vehicle 300 in which the imaging device 310 that captured the video is installed. The time information is information on the time when the video was captured. Furthermore, the acquisition unit 441 may acquire speed information transmitted from the recording device 320. The speed information includes at least information on the traveling speed of the vehicle 300, and may further include identification information and time information. The time information included in the speed information is information on the time when the traveling speed was recorded.

[0031] The adjustment unit 442 lowers the threshold of the person detection score in the peripheral region of a frame included in the video. That is, the threshold of the person detection score in the peripheral region of the frame becomes lower than the threshold of the person detection score in the central region, which is a region other than the peripheral region. Here, the central region is an area that is appropriately set depending on the purpose of person detection, and is a range of a predetermined size that includes the center of the frame. The center of the frame and the center of the central region may or may not coincide with each other.

[0032] The person detection unit 443 detects people from the video acquired by the acquisition unit 441. Specifically, the person detection unit 443 calculates a person detection score for each frame constituting the video acquired by the acquisition unit 441. The method of calculating the person detection score by the person detection unit 443 is similar to that of the person detection unit 140, and therefore description thereof will be omitted. Next, the person detection unit 443 determines whether the calculated person detection score is equal to or greater than the person detection score threshold 431. The person detection unit 443 calculates and determines the person detection score for each of multiple frames. The person detection score threshold 431 is a preset numerical value and is used when detecting people from frames constituting the video. The person detection score threshold 431 may be set to a different value depending on the area within the frame. Specifically, the person detection score threshold 431 is adjusted by the adjustment unit 442 in predetermined cases. The person detection unit 443 determines that a person is present in a location within a frame where the person detection score is equal to or greater than the person detection score threshold 431.

[0033] The masking unit 444 performs a masking process on an area in the frame that corresponds to a person detected by the person detection unit 443. Here, the masking process is image processing that is performed on the area so that the person cannot be identified, such as a solid color process or a filter process. The masking unit 444 may also perform a masking process on a part of the area that corresponds to the person in the frame (for example, a part that corresponds to the face).

[0034] FIG. 6 shows an example of a frame 10A included in a video acquired by the acquisition unit 441. The frame 10A shown in FIG. 6 shows a person 30A and a vehicle 40A traveling forward on a roadway 50A. In the example shown in FIG. 6, the adjustment unit 442 lowers the threshold 431 of the person detection score in the peripheral region R4. In FIG. 6, a central region R5 other than the peripheral region R4 is indicated by a dashed line. Next, the person detection unit 443 detects a person from the frame. Note that the threshold 431 of the person detection score in the peripheral region R4 is lowered compared to the threshold 431 of the person detection score in the central region R5, making it easier to detect a person in the peripheral region R4 than in the central region R5. In FIG. 6, a region R6 corresponding to the person 30A detected by the person detection unit 443 is indicated by a dashed line.

[0035] Next, a person detection method according to the second embodiment will be described with reference to FIG. 7 . First, the acquisition unit 441 acquires video transmitted from the image capture device 310 (step S201). Next, the adjustment unit 442 lowers the person detection score threshold 431 for the peripheral region R4 (step S202). Next, the person detection unit 443 detects a person from the frames constituting the video based on a predetermined person detection score threshold (step S203). Next, the masking unit 444 performs a masking process on the region corresponding to the person in the frame (step S204). Here, the person detection score threshold for the peripheral region R4 is lower than the person detection score threshold for the central region R5. Therefore, the person detection unit 443 can detect the person 30A more accurately in the peripheral region R4 than in the central region R5. This allows the person 30A to be accurately detected and masked in the peripheral region R4, while avoiding the inconvenience of erroneously detecting a person in the central region R5 and performing a masking process.

[0036] In this way, the human detection device 400 according to this embodiment detects people by lowering the threshold value of the human detection score in the area where it is highly likely that a person is included (peripheral area R4), and therefore can detect people with high accuracy. Furthermore, the threshold value of the human detection score cannot be lowered in the area where it is low likely that a person is included (central area R5), and therefore it is possible to avoid the inconvenience of erroneously detecting a person in this area R5. Therefore, the human detection device 400 according to this embodiment can detect people with high accuracy and suitably.

[0037] <Embodiment 3> Fig. 8 is a block diagram showing the configuration of a human detection device 600 according to embodiment 3. The human detection device 600 differs from the human detection device 400 shown in Fig. 5 in that it includes a control unit 640 instead of the control unit 440. The control unit 640 includes a roadway detection unit 641 and an adjustment unit 642 whose configurations differ from those of the control unit 440. Therefore, the configurations of the acquisition unit 441, human detection unit 443, and masking unit 444 of the control unit 640 overlap with those of embodiment 2, and therefore description thereof will be omitted where appropriate.

[0038] The control unit 640 includes an acquisition unit 441, a roadway detection unit 641 as roadway detection means, an adjustment unit 642, a person detection unit 443, and a masking unit 444. The acquisition unit 441 acquires video from the imaging device 310, and may also acquire speed information from the recording device 320. The person detection unit 443 detects people based on a threshold value of the person detection score for each of the multiple frames that make up the video acquired by the acquisition unit 441. The masking unit 444 performs masking processing on areas in the frames that correspond to people detected by the person detection unit 443.

[0039] The roadway detection unit 641 detects roadways from the video acquired by the acquisition unit 441 based on a predetermined roadway detection score threshold. Specifically, the roadway detection unit 641 detects roadways for each frame included in the video. The roadway detection unit 641 detects roadways using a trained roadway detection model (not shown). The roadway detection score is a score calculated by the roadway detection unit 641 for each region of a predetermined size within a frame using the roadway detection model. The roadway detection score is higher in regions where a roadway is likely to exist than in other regions. The predetermined roadway detection score threshold is a preset value and is used when detecting roadways from frames constituting the video. If the roadway detection score for a region within a frame is equal to or greater than the roadway detection score threshold, the roadway detection unit 641 determines that a roadway exists in the region. Here, a roadway refers to a portion of a road (excluding bicycle lanes) intended exclusively for vehicular traffic, including shoulders, side strips, and trackbeds. Furthermore, the method for calculating the roadway detection score is not limited to the above, and other existing techniques can be applied.

[0040] When the roadway detection unit 641 detects a roadway from a frame included in the video, the adjustment unit 642 lowers the threshold of the person detection score in an area other than the area corresponding to the roadway in the frame. In other words, the threshold of the person detection score in an area other than the area corresponding to the roadway in the frame becomes lower than the threshold of the person detection score in the area corresponding to the roadway.

[0041] FIG. 9 shows an example of a frame 10B included in the video acquired by the acquisition unit 441. Frame 10B shown in FIG. 9 includes a person 30B, a vehicle 40B traveling forward, and a roadway 50B on which the vehicle 40B is traveling. In the example shown in FIG. 9, the roadway detection unit 641 detects the roadway 50B. In FIG. 9, a region R7 corresponding to the roadway 50B detected by the roadway detection unit 641 is indicated by a dashed line. Next, the adjustment unit 642 lowers the person detection score threshold 431 in regions other than region R7. Next, the person detection unit 443 detects a person from the frame. Note that the person detection score threshold 431 in regions other than region R7 is lowered compared to the person detection score threshold 431 in region R7, making it easier to detect people in regions other than region R7 than in region R7. In FIG. 9, a region R8 corresponding to person 30B detected by the person detection unit 443 is indicated by a dashed line.

[0042] Next, a person detection method according to the third embodiment will be described with reference to FIG. 10 . First, the acquisition unit 441 acquires video transmitted from the image capture device 310 (step S301). Next, the roadway detection unit 641 detects a roadway from a frame included in the video based on a predetermined roadway detection score threshold (step S302). Next, the adjustment unit 642 lowers the person detection score threshold 431 for a region other than the region R7 corresponding to the roadway in the frame (step S303). Next, the person detection unit 443 detects a person from the frame constituting the video based on a predetermined person detection score threshold (step S304). Next, the masking unit 444 performs a masking process on the region corresponding to the person in the frame (step S305). Here, the person detection score threshold for the region other than the region R7 corresponding to the roadway is lower than the person detection score threshold for the region R7. Therefore, the person detection unit 443 can detect the person 30A more accurately in the region other than the region R7 corresponding to the roadway than in the region R7. This allows the person 30A to be accurately detected and masked in areas other than the area R7 corresponding to the roadway, and also avoids the inconvenience of erroneously detecting a person in the area R7 and performing masking.

[0043] In this way, the human detection device 600 according to this embodiment detects people by lowering the threshold of the human detection score in areas where it is highly likely that a person is included (areas other than area R7 corresponding to the roadway), and therefore can accurately detect people. Furthermore, because the threshold of the human detection score cannot be lowered in areas where it is low likely that a person is included (area R7 corresponding to the roadway), it is possible to avoid the inconvenience of erroneously detecting a person in this area R7. Therefore, the human detection device 600 according to this embodiment can accurately and suitably detect people.

[0044] Although the above-described embodiments have been described as hardware configurations, the present disclosure is not limited to such configurations. The processes described in Figures 3, 7, and 10 can also be realized by causing a CPU to execute a computer program.

[0045] In the above examples, the program can be stored and supplied to a computer using various types of non-transitory computer-readable media. Non-transitory computer-readable media include various types of tangible storage media. Examples of non-transitory computer-readable media include magnetic recording media (e.g., flexible disks, magnetic tapes, hard disk drives), magneto-optical recording media (e.g., magneto-optical disks), CD-ROMs (Read Only Memory), CD-Rs, CD-RWs, DVDs (Digital Versatile Discs), and semiconductor memories (e.g., mask ROMs, PROMs (Programmable ROMs), EPROMs (Erasable PROMs), flash ROMs, and RAMs (Random Access Memory)). The program may also be supplied to a computer by various types of transitory computer-readable media. Examples of transitory computer-readable media include electrical signals, optical signals, and electromagnetic waves. The temporary computer-readable medium can supply the program to the computer via a wired communication path such as an electric wire or an optical fiber, or via a wireless communication path.

[0046] The present disclosure is not limited to the above-described embodiments and may be modified as appropriate without departing from the spirit and scope of the present disclosure. Furthermore, the present disclosure may be implemented by appropriately combining the respective embodiments. Furthermore, the image capture device 310 of each vehicle 300 may be equipped with the functions of the object detection unit 120, adjustment unit 130, and person detection unit 140 of the person detection device 100. Similarly, the image capture device 310 of each vehicle 300 may be equipped with the functions of the adjustment unit 442, person detection unit 443, and masking unit 444 of the person detection device 400. Similarly, the image capture device 310 of each vehicle 300 may be equipped with the functions of the roadway detection unit 641, adjustment unit 642, person detection unit 443, and masking unit 444 of the person detection device 600. This allows the image capture device 310 of each vehicle 300 to individually perform person detection and masking processing.

[0047] Although the present invention has been described above with reference to the embodiments, the present invention is not limited to the above-described embodiments. Various modifications that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention.

[0048] 10, 10A, 10B Frame 20 Bicycle (two-wheeled vehicle) 30, 30A, 30B Person 40A, 40B Vehicle 50A, 50B Roadway 100, 400, 600 Person detection device 410 Memory 420 Communication unit 430 Storage unit 431 Threshold 432 Program 440, 640 Control unit 110, 441 Acquisition unit (acquisition means) 120 Object detection unit (object detection means) 641 Roadway detection unit (roadway detection means) 130, 442, 642 Adjustment unit (adjustment means) 140, 443 Person detection unit (person detection means) 444 Masking unit 200 Person detection system 300 Vehicle 310 Photographing device 311 Photographing unit 312 Communication unit 320 Recording device 321 Measuring unit 322 Communication unit 500 Network R1, R2, R3, R4, R5, R6, R7, R8 area

Claims

1. An acquisition means for acquiring the captured video; an object detection means for detecting a predetermined detection object from the video based on a predetermined object detection score threshold; an adjustment means for lowering a threshold value of a human detection score in a predetermined range of areas including the detection target in the video image to a value lower than the threshold value of the human detection score in other areas; a person detection means for detecting a person from the video based on the threshold value of the person detection score, Person detection device.

2. The detection target includes at least a two-wheeled vehicle. The human detection device according to claim 1 .

3. further comprising roadway detection means for detecting a roadway from the image; the adjustment means lowers the threshold value of the human detection score in an area other than the roadway in the video image to be lower than the threshold value of the human detection score in an area corresponding to the roadway; The human detection device according to claim 1 .

4. an imaging device installed in a vehicle and configured to capture an image of the surroundings of the vehicle; a person detection device capable of communicating with the photographing device, The person detection device an acquisition means for acquiring an image captured by the imaging device; an object detection means for detecting a predetermined detection object from the video based on a predetermined object detection score threshold; an adjustment means for lowering a threshold value of a human detection score in a predetermined range of areas including the detection target in the video image to a value lower than the threshold value of the human detection score in other areas; a person detection means for detecting a person from the video based on the threshold value of the person detection score, People detection system.

5. The person detection device further comprising roadway detection means for detecting a roadway from the image; the adjustment means lowers the threshold value of the human detection score in an area other than the roadway in the video image to be lower than the threshold value of the human detection score in an area corresponding to the roadway; The person detection system of claim 4 .

6. The computer Obtain the captured footage, Detecting a predetermined detection target from the video based on a predetermined target detection score threshold; lowering a threshold value of the human detection score in a predetermined range of areas including the detection target in the video image to be lower than the threshold value of the human detection score in other areas; detecting a person from the video based on the threshold value of the person detection score; Person detection methods.

7. The computer Further detecting a roadway from the image; lowering the threshold value of the human detection score in an area other than the roadway in the video to be lower than the threshold value of the human detection score in an area corresponding to the roadway; The person detection method according to claim 6 .

8. On the computer, A process of acquiring the captured video; detecting a predetermined detection target from the video based on a predetermined target detection score threshold; a process of lowering a threshold value of a person detection score in a predetermined range of areas including the detection target in the video image to a value lower than the threshold value of the person detection score in other areas; detecting a person from the video based on the threshold value of the person detection score; A person detection program that executes the following.

9. On the computer, Further detecting roadways from the video; a process of lowering the threshold value of the human detection score in an area other than the roadway in the video image to be lower than the threshold value of the human detection score in an area corresponding to the roadway; The person detection program according to claim 8 , wherein the program executes the following steps.