Detection device, detection system, detection method, and program

Abstract

To provide a detection device capable of improving accuracy in detecting a traffic accident.SOLUTION: A detection device includes an acoustic signal acquisition section, an accident sound detection section, an accident occurrence position estimation section, a captured image acquisition section, and an output section. The acoustic signal acquisition section acquires an acoustic signal collected by a sound collection device arranged along a road. The accident sound detection section detects accident sound from the acoustic signal. The accident occurrence position estimation section estimates position information of an accident occurrence position being a sound source position of the accident sound. The captured image acquisition section transmits the position information to an imaging device capable of capturing the accident occurrence position, and acquires a captured image corresponding to the position information from the imaging device. The output section generates output information including the position information and the captured image, so as to output them.SELECTED DRAWING: Figure 11

Description

[Technical field]

[0001] The present disclosure relates to a detection device, a detection system, a detection method, and a program. [Background technology]

[0002] A technology for detecting traffic accidents by acoustic signal processing is disclosed in Patent Document 1. The street light system described in Patent Document 1 detects a target sound indicating a traffic accident (hereinafter also referred to as an accident sound) and its occurrence position based on collected sound data. The street light system described in Patent Document 1 detects the accident sound by a sound recognition algorithm based on machine learning. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2020-087883 A Summary of the Invention [Problem to be solved by the invention]

[0004] It is difficult to detect accident sounds using sound recognition algorithms. The reason for this is that the collected sound data contains a lot of noise. For example, the street light system described in Patent Document 1 may erroneously detect sounds that are not caused by accidents as accident sounds. When an administrative office or the like that manages roads arranges for a response to an accident based on the detection result, if the administrative office arranges for a response to the accident without knowing that it is a false detection, it will result in unnecessary work. Therefore, in the detection of traffic accidents, it is desirable to further improve the detection accuracy.

[0005] An object of the present disclosure is to provide a detection device, a detection system, a detection method, and a program that can improve the accuracy of detecting traffic accidents. [Means for solving the problem]

[0006] A detection device according to one aspect of the present disclosure includes an acoustic signal acquisition unit that acquires an acoustic signal picked up by a sound collecting device installed along a road, an accident sound detection unit that detects an accident sound from the acoustic signal, an accident location estimation unit that estimates location information of the accident location which is the sound source location of the accident sound, an image acquisition unit that transmits location information to an imaging device capable of photographing the accident location and acquires an image captured from the imaging device according to the location information, and an output unit that generates and outputs output information including the location information and the captured image.

[0007] A detection system according to one aspect of the present disclosure is a traffic accident detection system comprising multiple microphone arrays and multiple photographing devices installed along a road, wherein each of the multiple microphone arrays includes multiple microphones and has a sound collection unit that collects acoustic signals, an accident sound detection unit that detects accident sounds from the acoustic signals, an accident location estimation unit that estimates position information of the accident location which is the sound source position of the accident sound, and a transmission unit that transmits position information to a photographing device capable of photographing the accident location, and each of the multiple photographing devices has a control unit that changes the photographing direction using the position information received from the transmission unit, a photographing unit that photographs an image, and an output unit that outputs the accident location and the photographed image.

[0008] In one aspect of the detection method of the present disclosure, a computer acquires an acoustic signal picked up by a sound collecting device installed along a road, detects an accident sound from the acoustic signal, estimates position information of the accident location which is the sound source position of the accident sound, transmits the position information to a camera capable of photographing the accident location, acquires an image captured from the camera device according to the position information, and generates and outputs output information including the position information and the captured image.

[0009] A program according to one aspect of the present disclosure causes a computer to execute the following processes: acquiring an acoustic signal picked up by a sound collecting device installed along a road; detecting an accident sound from the acoustic signal; estimating position information of the accident location, which is the sound source location of the accident sound; transmitting the position information to a camera capable of photographing the accident location; acquiring an image captured from the camera according to the position information; and generating and outputting output information including the position information and the captured image. Effect of the Invention

[0010] According to the present disclosure, it is possible to provide a detection device, a detection system, a detection method, and a program that improve the accuracy of detecting traffic accidents. [Brief description of the drawings]

[0011] [Figure 1] FIG. 1 is a conceptual diagram for explaining a detection system in the present disclosure. [Diagram 2] 1 is a block diagram showing an example of a configuration of a detection system according to the present disclosure. [Diagram 3] FIG. 4 is a diagram illustrating an example of output information generated by an output unit in the present disclosure. [Figure 4] 4 is a flowchart illustrating an example of an operation of the detection device according to the present disclosure. [Diagram 5] 5 is a flowchart illustrating an example of the operation of the imaging device according to the present disclosure. [Figure 6] 1 is a block diagram showing an example of a configuration of a detection system according to the present disclosure. [Figure 7] 1 is a block diagram showing an example of a configuration of a detection system according to the present disclosure. [Figure 8] 4 is a flowchart illustrating an example of an operation of the detection device according to the present disclosure. [Figure 9] 1 is a block diagram showing an example of a configuration of a detection system according to the present disclosure. [Figure 10] 4 is a flowchart illustrating an example of an operation of the detection device according to the present disclosure. [Figure 11]1 is a block diagram showing an example of a configuration of a detection device according to the present disclosure. [Figure 12] 4 is a flowchart illustrating an example of an operation of the detection device according to the present disclosure. [Figure 13] FIG. 2 is a diagram illustrating an example of a hardware configuration for executing control and processing in the present disclosure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the drawings. The embodiments described below are limited in a manner that is technically preferable for implementing the present disclosure, but the scope of the invention is not limited to the following. In all drawings used to explain the following embodiments, the same reference numerals are used for similar parts unless otherwise specified. In the following embodiments, repeated explanations of similar configurations and operations may be omitted.

[0013] [First embodiment] First, a detection system according to the first embodiment will be described with reference to the drawings. The detection system of this embodiment estimates the sound source position (accident occurrence position) where an accident has occurred based on a sound signal collected by a sound collection device arranged along a road. The detection system of this embodiment directs the photographing direction of a photographing device arranged along the road toward the estimated accident occurrence position.

[0014] (overview) Fig. 1 is a conceptual diagram for explaining an overview of a detection system in the present disclosure. Note that Fig. 1 is a diagram for explaining the overview of the detection system, and does not limit the shape or number of each component.

[0015] The detection system 1 includes a plurality of sound collection devices 11-n and a plurality of image capture devices 21-m (m and n are natural numbers). The sound collection devices 11-n and the image capture devices 21-m are installed along a road. The shapes and the number of the sound collection devices 11-n and the image capture devices 21-m are not limited to the example shown in FIG. 1. In the following description, when it is not necessary to distinguish between the sound collection devices 11-n, they will be referred to as sound collection devices 11. Similarly, when it is not necessary to distinguish between the image capture devices 21-m, they will be referred to as image capture devices 21. The detection system 1 also includes a detection device (not shown). For example, the detection device is included in the sound collection device 11 or the image capture device 21.

[0016] The sound collection device 11-n is a kind of electroacoustic transducer that converts acoustic vibrations into acoustic signals and collects sound. The sound collection device 11-n in this embodiment is realized by a microphone array consisting of multiple microphones, but may be realized by other electroacoustic transducers. Each sound collection device 11-n includes multiple microphones 111-l (l is a natural number). The sound collection device 11-n converts surrounding acoustic vibrations into acoustic signals using the multiple microphones 111-l (S1). The multiple microphones 111-l collect the converted acoustic signals. A detection device (not shown) analyzes the acoustic signals collected by the multiple microphones 111-l by acoustic signal processing. The detection device detects accident sounds from the collected acoustic signals. The detection device also estimates the accident occurrence position, which is the sound source of the accident sound. The detection device transmits position information of the estimated accident occurrence position to an image capture device 21-m that can capture the accident occurrence position (S2). The image capturing device 21-m, which has received the location information of the accident location, adjusts the image capturing angle and image capturing direction according to the received location information, and then captures the image (S3). The location information of the accident location and the captured image are output to an administrator terminal 30, which is a terminal device owned by an administration office that manages roads, for example.

[0017] (composition) 2 is a block diagram showing an example of the configuration of a detection system according to the present disclosure. The detection system 1 includes a sound collection device 11, a photographing device 21, and an administrator terminal 30. The sound collection device 11, the photographing device 21, and the administrator terminal 30 are connected to each other so as to be able to communicate with each other.

[0018] The sound collection device 11 has a sound collection unit 111, an acoustic signal acquisition unit 112, an accident sound detection unit 113, an accident occurrence position estimation unit 114, a photographed image acquisition unit 115, an output unit 116, and a storage unit 117. The acoustic signal acquisition unit 112, the accident sound detection unit 113, the accident occurrence position estimation unit 114, the photographed image acquisition unit 115, and the output unit 116 constitute a detection device 100. In this embodiment, the sound collection device 11 is configured to include the detection device 100, but this is not limited to the above. For example, the detection device 100 may be included in the photographing device 21. Alternatively, the detection device 100 may be realized by an independent information processing device. For example, the detection device 100 may be disposed in a server or a cloud connected to the sound collection device 11 and the photographing device 21 via a communication network. The detection device 100 may also be included in the administrator terminal 30.

[0019] The sound collection unit 111 is an example of a sound collection means for collecting surrounding acoustic signals. The sound collection unit 111 includes a plurality of microphones 111-1 to 1. It is preferable that each of the microphones 111-1 is an omnidirectional (omnidirectional) microphone, but this is not a limitation. The arrangement of each of the microphones 111-1 is not particularly limited as long as it allows sound source localization. Similarly, the number of microphones 111-1 is not particularly limited as long as it is one or more.

[0020] The acoustic signal acquiring unit 112 acquires an acoustic signal collected by the sound collecting unit 111. Specifically, the acoustic signal acquiring unit 112 acquires at least one of frequency, spectrum data, and waveform data.

[0021] The accident sound detection unit 113 detects an accident using the sound signal acquired by the sound signal acquisition unit 112. Specifically, the accident sound detection unit 113 detects an accident sound from the sound signal acquired by the sound signal acquisition unit 112. An accident sound refers to a sound signal that is estimated to be derived from a sound caused by an accident, among sound signals derived from abnormal sounds different from environmental sounds. An accident sound is also referred to as an accident sound signal. For example, the accident sound detection unit 113 detects an accident sound using the abnormal sound detection techniques listed below. The abnormal sound detection techniques listed below are examples and are not limited to these.

[0022] For example, the accident sound detection unit 113 detects an accident sound using a pre-created accident sound determination model. The accident sound determination model is stored in the storage unit 117. The accident sound determination model is machine-learned using normal sounds and accident sounds as training data. The training data is preferably both normal sounds and accident sounds, but may be either one of them. The accident sound detection unit 113 detects an accident sound from the picked-up acoustic signal by inputting the picked-up acoustic signal to the accident sound determination model.

[0023] For example, the accident sound detection unit 113 may detect an accident sound by comparing the sound signal of the accident sound stored in the storage unit 117 with the picked-up sound signal.

[0024] For example, the accident sound detection unit 113 may detect an accident sound by comparing a normal sound sound signal stored in the storage unit 117 with the picked-up sound signal.

[0025] The accident occurrence position estimation unit 114 estimates position information of the accident occurrence position, which is the sound source position of the accident sound. For example, the accident occurrence position estimation unit 114 estimates the position information of the accident occurrence position using the cross-correlation method disclosed in Non-Patent Document 1 (Non-Patent Document 1: Sato Tatsuki, Watanabe Seiji, "Comparison of Accuracy of Sound Source Position Estimation by Cross-Correlation Method and Application of Sound Source Localization", Proceedings of the 2018 National Conference of the Society of Industrial Applications Engineering, September 13, 2018).

[0026] In the cross-correlation method disclosed in Non-Patent Document 1, first, the accident occurrence position estimation unit 114 estimates the sound source direction of the acoustic signal collected by the sound collection unit 111 using a known sound source localization technique. For example, the known sound source localization technique includes a time delay method, a MUSIC (Multiple Signal Classification) method, a beamforming method, etc. The accident occurrence position estimation unit 114 estimates the sound source position, i.e., the accident occurrence position, using the sound source direction of each microphone. When using the following method, the sound collection device 11 includes three microphones 111-l (microphones 111-1, 111-2, and 111-3). The microphones 111-1, 111-2, and 111-3 are placed on the xy plane. The microphones 111-1 and 111-2 are placed equidistant from the origin on the x axis. Microphone 111-3 is installed on the y-axis at a distance equal to the distance from the origin to microphone 111-1. The sound source direction estimated by microphones 111-1 and 111-2 is defined as θ1, and the sound source direction estimated by microphones 111-1 and 111-3 is defined as θ2. Accident occurrence position estimation unit 114 calculates the equations of two straight lines A and B using the sound source directions θ1 and θ2. Straight line A is determined by the following equation 1. Straight line B is determined by the following equation 2.

[0027]

number

[0028]

number

[0029] The photographed image acquisition unit 115 transmits the position information estimated by the accident occurrence position estimation unit 114 to the photographing device 21-m capable of photographing the accident occurrence position. The photographed image acquisition unit 115 acquires the photographed image from the photographing device 21-m according to the position information. The photographed image acquisition unit 115 selects an arbitrary photographing device 21-m from the plurality of photographing devices 21 using the accident occurrence position information and the position information of the photographing device 21-m. The photographed image acquisition unit 115 transmits the position information to the photographing device 21-m capable of photographing the accident occurrence position. For example, the photographed image acquisition unit 115 transmits the position information to the photographing device 21-m closest to the accident occurrence position. The photographed image acquisition unit 115 may transmit the position information to the plurality of photographing devices 21-m capable of photographing the accident occurrence position according to the distance from the photographing device 21-m to the accident occurrence position and the focal length of the photographing device 21-m. The photographed image acquisition unit 115 acquires, from the photographing device 21-m, photographed images captured in accordance with the position information.

[0030] The output unit 116 generates output information including the position information estimated by the accident occurrence position estimation unit 114 and the captured image acquired by the captured image acquisition unit 115. The output unit 116 outputs the generated output information. The output unit 116 outputs the output information to the manager terminal 30 owned by the management station. Note that the output destination of the output information is not limited to the manager terminal 30.

[0031] FIG. 3 is a diagram showing an example of output information generated by the output unit 116. As shown in FIG. 3, the output information includes at least position information (D2) of the accident occurrence position and a captured image (D4) acquired by the captured image acquisition unit 115. FIG. 3 shows an example in which the captured image acquisition unit 115 acquires a plurality of captured images, but is not limited to this. When the captured image acquisition unit 115 acquires only a single captured image, the output information may include only one captured image (D4). The output information may also include information (D1) indicating that an accident has occurred and an audio file (D3) of the accident sound collected by the sound collection device 11.

[0032] The storage unit 117 stores information used in the processing of the detection device 100. When the detection device 100 is a part of the configuration of the sound collection device 11-n, the storage unit 117 stores the position information of each of the image capture devices 21-m. In addition, the storage unit 117 further stores at least one of an accident sound determination model, an acoustic signal of an accident sound, and an acoustic signal of a normal sound, depending on the method of detecting the accident sound. When the detection device 100 is realized by an independent information processing device, the storage unit 117 stores the position information of each of the sound collection devices 11-n and the position information of each of the image capture devices 21-m. The storage unit 117 further stores at least one of an accident sound determination model, an acoustic signal of an accident sound, and an acoustic signal of a normal sound, depending on the method of detecting the accident sound. When the detection device 100 is a part of the configuration of the sound collection device 11-n, the storage unit 117 stores the position information of each of the sound collection devices 11-n. The storage unit 117 further stores at least one of an accident sound determination model, an accident sound acoustic signal, and a normal sound acoustic signal, depending on the method for detecting an accident sound.

[0033] The photographing device 21 includes a photographing unit 211 and a control unit 212. For example, the photographing device 21 is realized by a photographing device such as a digital camera. For example, the photographing device 21 may be realized by a surveillance camera capable of controlling the photographing direction. The photographing device 21 is provided on a camera platform (not shown). The photographing device 21 may be integrated with the camera platform. The photographing unit 211 is an example of a photographing means for photographing still images or videos. The control unit 212 is an example of a control means for controlling the photographing unit 211. Specifically, the control unit 212 issues a photographing command to the photographing unit 211 and controls the photographing direction of the photographing unit 211. The control unit 212 controls the photographing direction by changing the angle of a camera platform (not shown) provided in the photographing device 21. When the control unit 212 receives position information from the accident occurrence position estimation unit 114, it changes the photographing direction according to the position information. For example, the control unit 212 adjusts the photographing direction so that the received position information is at the center of the photographed image. When the control unit 212 changes the shooting direction of the shooting unit 211, it outputs a shooting command to the shooting unit 211. The shooting unit 211 takes a picture according to the shooting command. The shooting unit 211 transmits the shot image to the shot image acquisition unit 115.

[0034] The administrator terminal 30 is a terminal device owned by a management office that manages roads. For example, the administrator terminal 30 is located in the management office. For example, the administrator terminal 30 is realized by an information processing device such as a PC (Personal Computer), a tablet terminal, or a smartphone. The administrator terminal 30 has a display unit for displaying received output information. In FIG. 2, one administrator terminal 30 is shown, but this is just one example. There may be one or more administrator terminals 30, and there may be multiple administrator terminals.

[0035] (Detection device operation) Next, an example of the operation of the detection device 100 in this embodiment will be described with reference to the drawings. Fig. 4 is a flowchart showing an example of the operation of the detection device in the present disclosure.

[0036] 4, first, the acoustic signal acquisition unit 112 acquires an acoustic signal collected by the sound collection unit 111 (step S11). Specifically, the acoustic signal acquisition unit 112 acquires an acoustic signal collected by a plurality of microphones 111-l included in the sound collection device 11.

[0037] Next, the accident sound detection unit 113 detects an accident sound from the sound signal acquired by the sound signal acquisition unit 112 (step S12). For example, the accident sound detection unit 113 detects an accident sound by using an accident sound determination model generated in advance.

[0038] Next, the accident occurrence position estimation unit 114 estimates position information of the accident occurrence position, which is the sound source position of the accident sound (step S13). The accident occurrence position estimation unit 114 estimates the sound source direction using a known sound source localization technique. The accident occurrence position estimation unit 114 estimates the sound source position, i.e., the position information of the accident occurrence position, using the estimated sound source direction by a known sound source position estimation technique.

[0039] Next, the photographed image acquiring unit 115 transmits the location information estimated by the accident occurrence location estimating unit 114 to the photographing device 21-m capable of photographing the accident occurrence location (step S14). Furthermore, the photographed image acquiring unit 115 acquires the photographed image photographed according to the accident occurrence location information from the photographing device 21-m (step S15).

[0040] Next, the output unit 116 generates output information including the position information estimated by the accident occurrence position estimation unit 114 and the captured image acquired by the captured image acquisition unit 115 (step S16). The output unit 116 outputs the generated output information (step S17). For example, the output unit 116 outputs the generated output information to the administrator terminal 30 owned by the management station. After completing the process of step S16, the detection device 100 ends the series of processes described above.

[0041] (Operation of the imaging device) Next, an example of the operation of the photographing device 21 in this embodiment will be described with reference to the drawings. Fig. 5 is a flow chart showing an example of the operation of the photographing device in this disclosure.

[0042] First, the control unit 212 changes the shooting direction using the received accident occurrence position information (step S21). For example, the control unit 212 adjusts the shooting direction so that the received position information is at the center of the shot image. After changing the shooting direction, the control unit 212 outputs a shooting command to the shooting unit 211.

[0043] Next, the photographing unit 211 performs photographing in accordance with the photographing command (step S22).

[0044] Next, the photographing section 211 transmits the photographed image to the photographed image acquiring section 115 (step S23). After completing the process of step S23, the photographing device 21 ends the series of processes described above.

[0045] As described above, the detection device of this embodiment includes an acoustic signal acquisition unit, an accident sound detection unit, an accident position estimation unit, a captured image acquisition unit, and an output unit. The acoustic signal acquisition unit acquires an acoustic signal collected by a sound collection device installed along a road. The accident sound detection unit detects an accident sound from the acoustic signal acquired by the acquisition unit. The accident position estimation unit specifies position information of the accident position, which is the sound source position of the accident sound. The captured image acquisition unit transmits the position information of the accident position to an imaging device capable of capturing an image of the accident position, and acquires an image captured according to the position information. The output unit outputs output information including the position information and the captured image.

[0046] The detection device of this embodiment, with the above configuration, can output not only the result of the acoustic signal processing but also the captured image captured according to the position information of the accident occurrence position. A general technology for detecting an accident by acoustic signal processing can detect the accident sound, but cannot grasp the scale of the accident. Furthermore, with general acoustic signal processing, there is a risk of falsely detecting the accident sound. This is due to the high difficulty of modeling the target sound and the inclusion of noise in the collected sound data. The output information generated by the detection device of this embodiment includes the captured image captured according to the position information of the accident occurrence position. The manager who checks the screen of the terminal device on which the output information is displayed can grasp the scale of the accident by looking at the captured image. Furthermore, the manager can determine whether the accident detected by the sound actually occurred or is a false detection by looking at the captured image. In other words, according to the detection device of this embodiment, it is possible to improve the detection accuracy of traffic accidents.

[0047] In one aspect of this embodiment, the photographed image acquisition unit transmits location information to a plurality of photographing devices capable of photographing the accident occurrence position. The photographed image acquisition unit receives a plurality of photographed images from the plurality of photographing devices. The output unit generates output information including the plurality of photographed images. According to the detection device of this aspect, the accident scene is photographed at a plurality of angles of view by the plurality of photographing devices. As a result, a blind spot area in a photographed image photographed by a certain photographing device may be included in a photographed image photographed by another photographing device. In other words, according to the detection device of this aspect, it is possible to reduce a blind spot area that may be included in a photographed image. In other words, according to the detection device of this embodiment, it is possible to improve the detection accuracy of a traffic accident.

[0048] In one aspect of this embodiment, the sound collection device is realized by a microphone array having multiple microphones. In order to estimate the accident occurrence position, it is necessary to use the acoustic signals collected by the multiple microphones. For example, when the sound collection device is realized by a microphone, the same acoustic vibration needs to be collected by the multiple microphones, so the installation interval of the sound collection device is narrowed. This may increase the number of sound collection devices to be installed. When the sound collection device is realized by a microphone array, the above-mentioned problem does not occur. This is because the microphone array has multiple microphones, so the accident occurrence position can be estimated by a single microphone array.

[0049] (Modification) In this embodiment, the detection device 100 (sound collection device 11) is configured to include the output unit 116, but the photographing device 21 may also include the output unit 116. FIG. 6 is a block diagram showing an example of the configuration of the detection system 1-1 in a modified example of the present disclosure. The sound collection device 11-1 includes a transmission unit 118 instead of the photographed image acquisition unit 115. The transmission unit 118 is an example of a transmission means for transmitting position information of the accident occurrence position estimated by the accident occurrence position estimation unit 114 to the control unit 212. The output unit 116 generates output information including the position information received from the sound collection device 11 and the photographed image photographed by the photographing device 21. The output unit 116 outputs the generated output information. In the configuration of FIG. 6, the output information is transmitted from the photographing device 21-1 to the administrator terminal 30 immediately after the photographed image photographed according to the position information is photographed. Therefore, according to the configuration of FIG. 6, the timing of outputting the photographed image to the administrator terminal 30 can be accelerated compared to the case where the photographed image is output to the administrator terminal 30 via the sound collection device 11.

[0050] In this embodiment, the photographed image acquisition unit 115 is configured to transmit the position information estimated by the accident occurrence position estimation unit 114 to the photographing device 21-m capable of photographing the accident occurrence position, but is not limited to this configuration. For example, the photographed image acquisition unit 115 may be configured to transmit a control command specifying the rotation angle of the tripod head of the photographing device 21-m to the photographing device 21-m capable of photographing the accident occurrence position. In this case, the photographed image acquisition unit 115 uses the accident occurrence position and the position of the photographing device 21-m to determine the rotation angle of the tripod head required for the photographing unit 211 of the photographing device 21-m to photograph the direction of the photographing position. For example, the photographed image acquisition unit 115 calculates the rotation angle in the left-right direction (pan direction) and the rotation angle in the up-down direction (tilt direction) required to photograph the photographing position, assuming that the normal angle of the photographing device 21-m (for example, the azimuth angle and elevation / depression angle of the tripod head when the photographing direction of the photographing device 21-m is oriented in a predetermined reference direction) is 0 degrees. The captured image acquisition unit 115 transmits a control command including the calculated rotation angle in the left-right direction (pan direction) and the up-down direction (tilt direction) to the image capture device 21-m. The control unit 212 of the image capture device 21-m that receives the control command changes the image capture direction by rotating the camera platform in accordance with the control command. With the above-mentioned configuration, the detection device 100 of this embodiment can be applied to an image capture device that is not equipped with a calculation device such as a CPU (Central Processing Unit) for controlling the image capture direction.

[0051] [Second embodiment] Next, a detection system and a detection device according to the present embodiment will be described with reference to the drawings. In the following description, the same parts as those in the first embodiment will be omitted as appropriate. The detection device of the present embodiment differs from the detection device of the first embodiment in that it performs accident detection by image processing in addition to accident detection processing by acoustic signal processing.

[0052] (composition) 7 is a block diagram showing an example of the configuration of a detection system according to the present disclosure. The detection system 2 includes a sound collection device 12, an image capture device 22, a detection device 200, and an administrator terminal 30. The administrator terminal 30 is similar to the administrator terminal 30 of the first embodiment, and therefore a description thereof will be omitted.

[0053] The sound collection device 12-n is a type of electroacoustic transducer that converts acoustic vibrations into acoustic signals and collects sound. As in the first embodiment, the sound collection device 12-n of this embodiment is realized by a microphone array, but is not limited to this. The sound collection device 12 includes a sound collection unit 121. The sound collection unit 121 collects surrounding acoustic signals. The sound collection unit 121 includes a plurality of microphones 121-1 to 1.

[0054] The photographing device 22 includes a photographing section 221 and a control section 222. The photographing section 221 and the control section 222 are similar to the photographing section 211 and the control section 212 in the first embodiment, and therefore a description thereof will be omitted.

[0055] The detection device 200 includes an acoustic signal acquisition unit 201, an accident sound detection unit 202, an accident occurrence position estimation unit 203, a photographed image acquisition unit 204, an accident image detection unit 205, a storage unit 206, and an output unit 207. For example, the detection device 200 is realized by an independent information processing device. The detection device 200 may be realized as a part of the configuration of the sound collection device 12, the photographing device 22, or the administrator terminal 30. The acoustic signal acquisition unit 201, the accident sound detection unit 202, and the accident occurrence position estimation unit 203 are similar to the acoustic signal acquisition unit 112, the accident sound detection unit 113, and the accident occurrence position estimation unit 114 in the first embodiment, and therefore a description thereof will be omitted.

[0056] The photographed image acquisition unit 204 transmits the position information estimated by the accident occurrence position estimation unit 203 to the photographing device 22-m capable of photographing the accident occurrence position. The photographed image acquisition unit 204 acquires the photographed image taken according to the position information from the photographing device 22-m. For example, the photographed image acquisition unit 204 transmits the position information to the photographing device 22-m closest to the accident occurrence position. For example, the photographed image acquisition unit 204 may transmit the position information to a plurality of photographing devices 22-m capable of photographing the accident occurrence position. The photographed image acquisition unit 204 acquires the photographed image taken according to the position information from the photographing device 22-m.

[0057] The accident image detection unit 205 is an example of an accident image detection means that detects an accident using a captured image taken according to the accident occurrence position. When the captured image acquisition unit 204 acquires a plurality of captured images, the accident image detection unit 205 detects an accident for each captured image. For example, the accident image detection unit 205 detects an accident by the method described below. Note that the following method is an example and is not limited to this.

[0058] For example, the accident image detection unit 205 detects an accident using a pre-created accident image judgment model. The accident image judgment model is stored in the storage unit 206. The accident image judgment model is machine-learned using images at normal times and images at the time of an accident as training data. The training data is preferably both images at normal times and images at the time of an accident, but may be either one of them. The accident image detection unit 205 detects an accident from a captured image by inputting the captured image into the accident image judgment model. For example, the accident image detection unit 205 may compare an image at the time of an accident stored in the storage unit 206 with the captured image, and detect an accident according to the degree of similarity.

[0059] For example, the accident image detection unit 205 may detect an accident by comparing a captured image with an image stored in the storage unit 206 under normal circumstances.

[0060] For example, the accident image detection unit 205 may detect a traffic abnormality that hinders traffic flow as an accident. First, the accident image detection unit 205 detects a vehicle and its state from a captured image. If there is a vehicle whose state is different from the normal state, the accident image detection unit 205 determines that a traffic abnormality has occurred.

[0061] The storage unit 206 stores information used in the processing of the detection device 200. When the detection device 200 is a part of the configuration of the sound collection device 12-n, the storage unit 206 stores the position information of each of the image pickup devices 22-m. In addition, the storage unit 206 further stores at least one of an accident sound determination model, an acoustic signal of an accident sound, and an acoustic signal of a normal sound, depending on the method of detecting the accident sound. When the detection device 200 is realized by an independent information processing device, the storage unit 206 stores the position information of each of the sound collection devices 12-n and the position information of each of the image pickup devices 22-m. The storage unit 206 further stores at least one of an accident sound determination model, an acoustic signal of an accident sound, and an acoustic signal of a normal sound, depending on the method of detecting the accident sound. When the detection device 200 is a part of the configuration of the sound collection device 12-n, the storage unit 206 stores the position information of each of the sound collection devices 12-n. The storage unit 206 further stores at least one of an accident sound determination model, an acoustic signal of an accident sound, and an acoustic signal of a normal sound, depending on the method for detecting an accident sound.

[0062] When an accident is detected by the accident image detection unit 205, the output unit 207 generates output information including the position information estimated by the accident occurrence position estimation unit 203 and the captured image captured by the imaging device 22. When the captured image acquisition unit 204 acquires a plurality of captured images, the output unit 207 generates output information when an accident is detected in at least one of the plurality of captured images. The output unit 207 outputs the generated output information.

[0063] (Detection device operation) Next, the operation of the detection device 200 according to this embodiment will be described with reference to Fig. 8. Note that the operation of the photographing device 22 is similar to the operation shown in Fig. 5, and therefore the description will be omitted.

[0064] First, the acoustic signal acquisition unit 201 acquires an acoustic signal collected by the sound collection unit 121 (step S201). Specifically, the acoustic signal acquisition unit 201 acquires an acoustic signal collected by a plurality of microphones 121-l included in the sound collection device 12.

[0065] Next, the accident sound detection unit 202 detects an accident sound from the sound signal acquired by the sound signal acquisition unit 201 (step S202). For example, the accident sound detection unit 202 detects an accident sound by using an accident sound determination model created in advance.

[0066] Next, the accident occurrence position estimation unit 203 estimates position information of the accident occurrence position, which is the sound source position of the accident sound (step S203). The accident occurrence position estimation unit 203 estimates the sound source direction using a known sound source localization technique. The accident occurrence position estimation unit 203 estimates the sound source position, i.e., the position information of the accident occurrence position, using the estimated sound source direction by a known sound source position estimation technique.

[0067] Next, the photographed image acquiring unit 204 transmits the position information estimated by the accident occurrence position estimating unit 203 to the photographing device 22-m capable of photographing the accident occurrence position (step S204).

[0068] Next, the photographed image acquiring unit 204 acquires, from the photographing device 22-m, a photographed image captured in accordance with the accident occurrence position information (step S205).

[0069] Next, the accident image detection unit 205 detects an accident using the captured image captured according to the accident occurrence position (step S206). For example, the accident image detection unit 205 detects an accident using an accident image determination model created in advance.

[0070] Next, the output unit 207 judges whether or not an accident is detected by the accident image detection unit 205 (step S207). If an accident is not detected by the accident image detection unit 205 (No in step S207), the output unit 207 ends the process. If an accident is detected by the accident image detection unit 205 (Yes in step S207), the output unit 207 generates output information including the position information estimated by the accident occurrence position estimation unit 203 and the photographed image acquired by the photographed image acquisition unit 204 (step S208). The output unit 207 outputs the generated output information (step S209). For example, the output unit 207 outputs the generated output information to the administrator terminal 30 owned by the management station. When the detection device 100 finishes the process of step S208, it ends the series of processes described above.

[0071] The detection device of this embodiment is configured as described above. The detection device of this embodiment provides the following effects in addition to the effects described in the first embodiment.

[0072] The detection device of this embodiment further includes an accident image detection unit that detects an accident using the captured image acquired from the imaging device. In addition, the output unit of the detection device of this embodiment generates output information including the location information estimated by the accident location estimation unit and the captured image captured by the imaging device when an accident is detected by the accident image detection unit.

[0073] As described above, in general acoustic signal processing, there is a risk of erroneously detecting accident sounds. This is due to the high difficulty of modeling the target sound, the inclusion of noise in the collected sound data, and the like. After an accident sound is detected by acoustic signal processing, the detection device of this embodiment detects an accident again using a captured image. Even if the accident sound detection unit detects something other than an accident sound, the detection device of this embodiment does not generate and output output information unless an accident is detected by the accident image detection unit. Therefore, according to the detection device of this embodiment, it is possible to improve the detection accuracy of traffic accidents.

[0074] The captured image acquisition unit included in the detection device of one aspect of this embodiment transmits location information to the multiple image capture devices. The captured image acquisition unit acquires multiple captured images from the multiple image capture devices. In addition, the output unit generates output information including location information and the multiple captured images when an accident is detected in at least one of the multiple captured images. The output unit outputs the generated output information. For example, if an area where an accident occurred is a blind spot in an image captured by a certain image capture device, the accident cannot be detected from the image captured by that image capture device. However, the blind spot may not be a blind spot area in an image captured by another image capture device. Therefore, it is possible to detect an accident from an image captured by another image capture device. The detection device of this aspect generates output information when an accident is detected in at least one of the multiple captured images. In other words, the detection device of this aspect is less susceptible to the influence of the blind spot area, making it easier to detect an accident. In other words, the detection device of this embodiment can improve the detection accuracy of traffic accidents.

[0075] (Modification) The detection system 2 of this embodiment may be realized by the configuration of the detection system 2-1 shown in FIG. 9. The transmission unit 124 included in the sound collection device 12-1 is the same as the transmission unit 118 shown in the modified example of the first embodiment. The storage unit 206-1 stores information used for processing by the accident sound detection unit 202 and the accident occurrence position estimation unit 203 among the information stored in the storage unit 206. The storage unit 206-2 stores information used for processing by the accident image detection unit 205 among the information stored in the storage unit 206. The output unit 207 generates output information including the position information received from the sound collection device 12-1 and the captured image captured by the imaging device 22-1. In the configuration of FIG. 9, the output information is transmitted from the imaging device 22-1 to the administrator terminal 30 immediately after the captured image is captured according to the position information. Therefore, according to the configuration of FIG. 9, the timing of outputting the captured image to the administrator terminal 30 can be accelerated compared to the case where the captured image is output to the administrator terminal 30 via the detection device 200.

[0076] In the detection system 2 of this embodiment, the multiple sound collection devices 12-n are realized by a microphone array, but may be realized by microphones. The multiple microphones are arranged along the road. When the multiple sound collection devices 12 are realized by microphones, the accident occurrence position estimation unit 203 estimates position information of the accident occurrence position, for example, using the principle of triangulation. For example, according to the method disclosed in Patent Document 1 (JP Patent Publication No. 2020-087883), the position information of the accident occurrence position can be estimated using the principle of triangulation.

[0077] The detection device 200 of this embodiment performs the process of detecting an accident by image processing after the process of detecting an accident by sound signal processing, but this order may be reversed. The operation of the detection device 200 when the order of the process of detecting an accident by sound signal processing and the process of detecting an accident by image processing is reversed will be described with reference to FIG. 10. FIG. 10 is an example of the operation of the detection device in this modified example. In FIG. 10, first, the photographed image acquisition unit 204 acquires a photographed image photographed by the photographing device 22 (step S211). Next, the accident image detection unit 205 detects an accident using the photographed image (step S212). Next, the accident image detection unit 205 estimates the accident occurrence position using the photographed image (step S213). Next, the acoustic signal acquisition unit 201 transmits the accident occurrence position information to the sound collection device 12 capable of collecting the sound at the accident occurrence position (step S214). The acoustic signal acquisition unit 201 acquires the acoustic signal collected by the sound collection device 12 (step S215). In step S215, the sound signal acquisition unit 201 acquires sound signals acquired in a specific time period starting from the time when the accident was detected in step S212. For example, the sound signal acquisition unit 201 acquires sound signals in a time period going back by a time period predetermined by a user from the time when the accident was detected in step S212. Next, the accident sound detection unit 202 detects an accident sound from the acquired sound signals (step S216). If an accident sound is not detected (No in step S217), the output unit 207 ends the process. If an accident sound is detected (Yes in step S217), the output unit 207 generates output information including the accident occurrence position and the captured image (step S218). The output unit 207 outputs the generated output information (step S219).

[0078] [Third embodiment]

[0079] (composition) Next, the configuration of the detection system 3 according to this embodiment will be described with reference to the drawings. The detection system 3 has a simplified configuration of the detection device according to each embodiment. In the following description, the description of the same parts as those in the first or second embodiment will be omitted as appropriate.

[0080] 11 is a block diagram showing an example of the configuration of a detection device according to the present disclosure. The detection device 100 includes a sound signal acquisition unit 101, an accident sound detection unit 102, an accident occurrence position estimation unit 103, a captured image acquisition unit 104, and an output unit 105.

[0081] The sound signal acquiring unit 101 is an example of a sound signal acquiring means that acquires sound signals collected by sound collecting devices installed along the road.

[0082] The accident sound detection unit 102 is an example of an accident sound detection means that detects an accident sound from the acoustic signal acquired by the acoustic signal acquisition unit 101 .

[0083] The accident occurrence position estimating unit 103 is an example of an accident occurrence position estimating means that estimates position information of the accident occurrence position, which is the sound source position of the accident sound.

[0084] The captured image acquisition unit 104 is an example of an image acquisition means that transmits location information estimated by the accident location estimation unit 103 to an image capture device capable of capturing the location of the accident, and acquires an image captured according to the location information from the image capture device.

[0085] The output unit 105 is an example of an output means that outputs output information including position information and a captured image.

[0086] (operation) Next, an example of the operation of the detection device 100 according to the present disclosure will be described with reference to FIG.

[0087] First, the sound signal acquisition unit 101 acquires a sound signal collected by a sound collection device installed along a road (step S101).

[0088] Next, the accident sound detection unit 102 detects an accident sound from the acoustic signal acquired by the acoustic signal acquisition unit 101 (step S102).

[0089] Next, the accident occurrence position estimating unit 103 estimates position information of the accident occurrence position, which is the sound source position of the accident sound (step S103).

[0090] Next, the photographed image acquiring unit 104 transmits the position information estimated by the accident occurrence position estimating unit 103 to an imaging device capable of capturing an image of the accident occurrence position (step S104).

[0091] Next, the photographed image acquiring unit 104 acquires the photographed image captured according to the position information from the photographing device (step S105).

[0092] Next, the output unit 105 outputs the output information including the position information and the captured image (step S106).

[0093] According to the detection device of this embodiment, it is possible to obtain the same advantageous effects as the detection device of the first embodiment.

[0094] [Hardware configuration] The functions of each of the components in each of the embodiments of the present disclosure described above can be realized not only by hardware but also by a computer device or firmware based on program control.

[0095] Fig. 13 is a diagram showing an example of a hardware configuration of a detection device according to the present disclosure, which is realized by a computer device 90 including a processor. The detection device of each embodiment is realized by the computer device 90. As shown in Fig. 13, the computer device 90 includes a CPU (Central Processing Unit) 91, a memory 92, a storage device 93 such as a hard disk for storing programs, an input / output interface 94 for connecting input devices and output devices, and a communication interface 95 for connecting to a network.

[0096] The CPU 91 operates an operating system to control the prediction device of the present disclosure. For example, the CPU 91 reads out programs and data from a storage medium attached to a drive device or the like into the memory 92. The CPU 91 also functions as a part of the acoustic signal acquisition unit 101, the accident sound detection unit 102, the accident occurrence position estimation unit 103, the photographed image acquisition unit 104, and the output unit 105 in the detection device 100 of the present disclosure, and executes processes or commands based on the programs.

[0097] The storage device 93 is, for example, an optical disk, a flexible disk, a magneto-optical disk, an external hard disk, or a semiconductor memory. A part of the storage device is a storage medium that is a non-volatile storage device, and the program is recorded therein. The program may also be downloaded from an external computer (not shown) connected to a communication network.

[0098] The input device connected to the input / output interface 94 is realized by, for example, a mouse, a keyboard, etc., and is used for input operations. Similarly, the output device connected to the input / output interface 94 is realized by, for example, a display, etc., and is used for displaying and checking output results.

[0099] Although the present disclosure has been described above with reference to each embodiment, the present disclosure is not limited to the above embodiment. Various changes that a person skilled in the art can understand within the scope of the present disclosure can be made to the configuration and details of the present disclosure. For example, although multiple operations are described in order in the form of a flowchart, the order of description does not limit the order in which the multiple operations are performed. Therefore, when implementing each embodiment, the order of the multiple operations can be changed to the extent that the content is not affected.

[0100] Some or all of the above embodiments can be described as follows:

[0101] (Appendix 1) an acoustic signal acquisition means for acquiring an acoustic signal collected by a sound collection device installed along a road; an accident sound detection means for detecting an accident sound from the acoustic signal; an accident occurrence position estimation means for estimating position information of an accident occurrence position, which is a sound source position of the accident sound; a photographed image acquiring means for transmitting the location information to a photographing device capable of photographing the accident occurrence location and acquiring a photographed image from the photographing device in accordance with the location information; an output means for outputting output information including the position information and the photographed image; A detection device comprising: (Appendix 2) 2. The detection device of claim 1, wherein the sound collection device is a microphone array having a plurality of microphones. (Appendix 3) The photographed image acquisition means includes: transmitting the position information to a plurality of the photographing devices; acquiring a plurality of the photographed images from the plurality of photographing devices; The output means includes: 3. The detection device of claim 2, further comprising: an output unit that outputs information including the location information and the plurality of captured images. (Appendix 4) The vehicle further includes an accident image detection means for detecting an accident using the acquired photographic image, The output means includes: 3. The detection device according to claim 2, wherein when an accident is detected by the accident image detection means, output information including the location information and the captured image is output. (Appendix 5) The photographed image acquisition means includes: transmitting the position information to a plurality of the photographing devices; acquiring a plurality of the photographed images from the plurality of photographing devices; The output means includes: The detection device of claim 4, further comprising: a detection unit configured to output output information including the location information and the plurality of captured images when an accident is detected in at least one of the plurality of captured images. (Appendix 6) A detection device according to any one of claims 1 to 5; A plurality of the sound collecting devices are installed along a road; a plurality of the photographing devices installed along a road; The sound collecting device is a sound collection means for collecting the acoustic signal, The imaging device is The camera includes a photographing means for photographing an image and a control means for controlling a photographing direction of the photographing means, The sound collection means is Transmitting the collected acoustic signal to the acoustic signal acquisition means; The control means controlling the photographing direction in accordance with the position information received from the accident occurrence position estimation; The photographing means is A detection system that transmits the captured image to the captured image acquisition means. (Appendix 7) A traffic accident detection system including a plurality of microphone arrays and a plurality of image capturing devices installed along a road, Each of the plurality of microphone arrays comprises: A sound collection means including a plurality of microphones for collecting an acoustic signal; an accident sound detection means for detecting an accident sound from the acoustic signal; an accident occurrence position estimation means for estimating position information of an accident occurrence position, which is a sound source position of the accident sound; a transmitting means for transmitting the location information to a photographing device capable of photographing the location where the accident occurred; having Each of the plurality of imaging devices is a control means for changing an image capturing direction using the position information received from the transmission means; An image capturing means for capturing an image; an output means for outputting the accident occurrence position and the photographed image photographed by the photographing means; having Detection system. (Appendix 8) Each of the plurality of imaging devices is The vehicle further includes an accident image detection unit that detects an accident using the captured image, The output means includes: 8. The detection system according to claim 7, wherein when an accident is detected by the accident image detection means, output information including the location information and the captured image is output. (Appendix 9) 3. The detection device of claim 2, wherein the output information further includes an audio file of the accident sound picked up by the multiple microphones. (Appendix 10) The computer Acquire an acoustic signal collected by a sound collection device installed along a road, Detecting an accident sound from the acoustic signal; Estimating position information of an accident occurrence position, which is a sound source position of the accident sound; transmitting the location information to an image capturing device capable of capturing an image of the location where the accident occurred; Acquire a photographed image photographed according to the location information from the photographing device; and outputting output information including the position information and the captured image. (Appendix 11) The detection method of claim 10, wherein the sound collection device is a microphone array having a plurality of microphones. (Appendix 12) The computer In transmitting the position information, the position information is transmitted to a plurality of the photographing devices; In acquiring the captured images, a plurality of the captured images are acquired from the plurality of image capturing devices; 12. The detection method according to claim 10, wherein, in the output, output information including the position information and the plurality of captured images is generated and output. (Appendix 13) The computer further 12. The detection method according to claim 10 or 11, further comprising detecting an accident using the acquired captured image. (Appendix 14) The computer In transmitting the position information, the position information is transmitted to a plurality of the photographing devices; In acquiring the captured images, a plurality of the captured images are acquired from the plurality of image capturing devices; The detection method of claim 13, wherein, in the output processing, when an accident is detected in at least one of the plurality of captured images, output information including the location information and the plurality of captured images is generated and output. (Appendix 15) 15. The detection method according to any one of claims 11 to 14, wherein the output information further includes an audio file of the accident sound picked up by the multiple microphones. (Appendix 16) On the computer, A process of acquiring an acoustic signal collected by a sound collecting device installed along a road; A process of detecting an accident sound from the acoustic signal; A process of estimating position information of an accident occurrence position, which is a sound source position of the accident sound; a process of transmitting the location information to an image capturing device capable of capturing an image of the location where the accident occurred; A process of acquiring a photographed image photographed according to the position information from the photographing device; generating and outputting output information including the position information and the photographed image; A program to execute. (Appendix 17) The program described in Appendix 10, wherein the sound collection device is a microphone array having a plurality of microphones. (Appendix 18) In the process of transmitting the position information, the position information is transmitted to a plurality of the photographing devices; In the process of acquiring the captured images, a plurality of the captured images are acquired from the plurality of image capturing devices; 18. The program according to claim 16, wherein in the outputting process, output information including the position information and the plurality of captured images is generated and output. (Appendix 19) The computer, in addition, 18. The program according to claim 16 or 17, which causes a process of detecting an accident to be executed using the acquired captured image. (Appendix 20) In the process of transmitting the position information, the position information is transmitted to a plurality of the photographing devices; In the process of acquiring the captured images, a plurality of the captured images are acquired from the plurality of image capturing devices; The program described in Appendix 19, wherein, in the output process, when an accident is detected in at least one of the plurality of captured images, output information including the location information and the plurality of captured images is generated and output. (Appendix 21) 21. The detection method according to any one of claims 17 to 20, wherein the output information further includes an audio file of the accident sound picked up by the multiple microphones. [Explanation of symbols]

[0102] 1, 2 Detection system 11, 12 Sound recording device 21, 22 Imaging device 30 Administrator terminal 100, 200 Detection device 111, 121 Sound pickup section 101, 112, 201 Acoustic signal acquisition unit 102, 113, 202 Accident sound detection unit 103, 114, 203 Accident location estimation section 104, 115, 204 Image acquisition unit 205 Accident image detection unit 105, 116, 207 Output section 117, 206 Memory section 211, 221 Filming Department 212, 222 Control unit 90 Computer Equipment 91 CPU(Central Processing Unit) 92 Memory 93 Storage device 94 Input / Output Interface 95 Communication Interface

Claims

1. an acoustic signal acquisition means for acquiring an acoustic signal collected by a sound collection device installed along a road; an accident sound detection means for detecting an accident sound from the acoustic signal; an accident occurrence position estimation means for estimating position information of an accident occurrence position, which is a sound source position of the accident sound; a photographed image acquiring means for transmitting the location information to a photographing device capable of photographing the accident occurrence location and acquiring a photographed image from the photographing device in accordance with the location information; an output means for outputting output information including the position information and the photographed image; A detection device comprising:

2. The detection device according to claim 1 , wherein the sound pickup device is a microphone array having a plurality of microphones.

3. The photographed image acquisition means includes: transmitting the position information to a plurality of the photographing devices; acquiring a plurality of the photographed images from the plurality of photographing devices; The output means includes: The detection device according to claim 2 , further comprising an output unit configured to output information including the position information and the plurality of captured images.

4. The vehicle further includes an accident image detection means for detecting an accident using the acquired photographic image, The output means includes: The detection device according to claim 2 , wherein when an accident is detected by said accident image detection means, output information including said location information and said photographed image is output.

5. The photographed image acquisition means includes: transmitting the position information to a plurality of the photographing devices; acquiring a plurality of the photographed images from the plurality of photographing devices; The output means includes: The detection device according to claim 4 , further comprising: an output unit configured to output information including the position information and the plurality of photographed images when an accident is detected in at least one of the plurality of photographed images.

6. A detection device according to any one of claims 1 to 5; A plurality of the sound collecting devices are installed along a road; a plurality of the photographing devices installed along a road; The sound collecting device is a sound collection means for collecting the acoustic signal, The imaging device is The camera includes an image capturing means for capturing the captured image, and a control means for controlling the image capturing direction of the image capturing means, The sound collection means is Transmitting the collected acoustic signal to the acoustic signal acquisition means; The control means controlling the photographing direction in accordance with the position information received from the accident occurrence position estimating means; The photographing means is taking the photographed image in response to control by the control means; A detection system that transmits the captured image to the captured image acquisition means.

7. A traffic accident detection system including a plurality of microphone arrays and a plurality of image capturing devices installed along a road, Each of the plurality of microphone arrays comprises: A sound collection means including a plurality of microphones for collecting an acoustic signal; an accident sound detection means for detecting an accident sound from the acoustic signal; an accident occurrence position estimation means for estimating position information of an accident occurrence position, which is a sound source position of the accident sound; a transmitting means for transmitting the location information to a photographing device capable of photographing the location where the accident occurred, Each of the plurality of imaging devices includes: a control means for changing an image capturing direction using the position information received from the transmission means; An image capturing means for capturing an image; an output means for outputting the accident occurrence position and the photographed image; having Detection system.

8. Each of the plurality of imaging devices includes: The vehicle further includes an accident image detection means for detecting an accident using the captured image, The output means includes: The detection system according to claim 7 , wherein when an accident is detected by said accident image detection means, output information including said location information and said photographed image is output.

9. The computer Acquire an acoustic signal collected by a sound collection device installed along a road, Detecting an accident sound from the acoustic signal; Estimating position information of an accident occurrence position, which is a sound source position of the accident sound; transmitting the location information to an image capturing device capable of capturing an image of the location where the accident occurred; Acquire a photographed image photographed according to the location information from the photographing device; and outputting output information including the position information and the captured image.

10. On the computer, A process of acquiring an acoustic signal collected by a sound collecting device installed along a road; A process of detecting an accident sound from the acoustic signal; A process of estimating position information of an accident occurrence position, which is a sound source position of the accident sound; a process of transmitting the location information to an image capturing device capable of capturing an image of the location where the accident occurred; A process of acquiring a photographed image photographed according to the position information from the photographing device; outputting output information including the position information and the captured image; A program to execute.

Images