Object detection system and object detection method

Abstract

To provide an object detection system having a plurality of cameras and enabling a monitoring target to be appropriately monitored by the cameras, and to provide an object detection method.SOLUTION: An object detection system 100 comprises: a plurality of cameras 120 in which a shooting space partially overlap each other; a detection part 112 that detects an object from a video image of the camera 120; a determination part 113 that determines whether or not the object is a monitoring target based on a state of the object; an identification part 114 that associates a spatial coordinate of the shooting spaces of the plurality of cameras 120 with each other and identifies a position of the object detected from the video image of one camera 120 in the shooting space of the other cameras 120; and a designation part 115 that designates the object detected from the video image of the other camera 120 as a monitoring target when it is determined that the object detected from the video image of the one camera 120 is the monitoring target and it is not determined that the object detected from the video image of the other camera 120 is the monitoring target at the position of the shooting space of the other camera 120 of the object.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an object detection system and an object detection method. [Background technology]

[0002] For example, if a pedestrian detected from video footage taken at an intersection by multiple cameras is not paying attention to his or her surroundings, it would be desirable to develop a monitoring system that notifies vehicles passing through the intersection of this fact. For example, Patent Document 1 describes a monitoring system that identifies objects from video footage taken by multiple cameras. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] International Publication No. 2017 / 163282 Summary of the Invention [Problem to be solved by the invention]

[0004] However, the shooting directions of multiple cameras are often different, and the same pedestrian may be judged as a target for surveillance in the footage captured by one camera, but not in the footage captured by another camera.

[0005] An object of the present invention is to provide an object detection system and an object detection method that can appropriately monitor a monitoring target using multiple cameras. [Means for solving the problem]

[0006] The object detection system of the present invention is an object detection system that includes a plurality of cameras, and at least a portion of the shooting spaces of the plurality of cameras overlap with each other, and includes: a detection unit that detects an object from images captured by the plurality of cameras; a judgment unit that judges whether the object is a monitoring target based on the state of the object detected by the detection unit; an identification unit that associates the spatial coordinates of the respective shooting spaces of the plurality of cameras with each other and identifies the position of an object detected from an image captured by one of the plurality of cameras in the shooting space of the other cameras among the plurality of cameras based on the position of the object in the shooting space of the one camera; and a designation unit that, if the object detected from the image captured by the one camera is judged to be a monitoring target, designates the object detected from the image captured by the other camera as a monitoring target at a position in the shooting space of the other camera that corresponds to the position of the object judged to be a monitoring target in the shooting space of the one camera.

[0007] The object detection method of the present invention is a method that executes the following steps: a detection step in which a computer detects an object from images captured by multiple cameras whose shooting spaces overlap at least partially; a determination step in which a computer determines whether the object is a surveillance target based on the state of the object detected in the detection step; an identification step in which spatial coordinates of the shooting spaces of the multiple cameras are associated with each other, and, based on the position of the object detected from the image captured by one of the multiple cameras in the shooting space of the one camera, a determination step in which a computer designates the object detected from the image captured by the other camera as a surveillance target if the object detected from the image captured by the one camera is determined to be a surveillance target at a position in the shooting space of the other camera that corresponds to the position of the object determined to be a surveillance target in the shooting space of the one camera. [Effects of the Invention]

[0008] According to the present invention, it is possible to provide an object detection system and an object detection method that can appropriately monitor a monitoring target using multiple cameras. [Brief explanation of the drawings]

[0009] [Figure 1] 1 is a block diagram showing an example of the configuration of an object detection system according to a first embodiment. [Figure 2] FIG. 2 is a top view showing an example of the arrangement of cameras according to the first embodiment. [Figure 3] 3 is a flowchart illustrating an example of an object detection method according to the first embodiment. [Figure 4] 3 is a flowchart illustrating an example of an object detection method according to the first embodiment. [Figure 5] FIG. 10 is a block diagram showing an example of the configuration of an object detection system according to a second embodiment. [Figure 6] 10 is a flowchart illustrating an example of an object detection method according to the second embodiment. [Figure 7] FIG. 10 is a block diagram showing an example of the configuration of an object detection system according to a third embodiment. [Figure 8] 11 is a flowchart illustrating an example of an object detection method according to a third embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0010] Embodiment 1 Hereinafter, an embodiment of the present invention will be described with reference to the drawings. 1 shows an example of the configuration of an object detection system 100 according to the first embodiment of the present invention. As shown in FIG. 1, the object detection system 100 includes a control unit 110, a first camera 121, a second camera 122, a third camera 123, a fourth camera 124, and a communication unit 130.

[0011] The object detection system 100 shown in FIG. 1 is an example of a system that monitors one intersection and is configured to include four cameras. The object detection system 100 may be a system that monitors multiple intersections in cooperation with one another, and FIG. 1 may represent a partial configuration thereof. The control unit 110 may be configured to control the cameras that monitor one intersection, or may be configured to control cameras installed at multiple intersections. The first camera 121, the second camera 122, the third camera 123, and the fourth camera 124 can communicate with the control unit 110 via a network (not shown). The number of cameras included in the object detection system 100 is not limited to four per intersection, as long as there are two or more cameras. Hereinafter, when there is no need to distinguish between the first camera 121, the second camera 122, the third camera 123, and the fourth camera 124, they will be simply referred to as "camera 120" or "multiple cameras 120." The object detection system 100 detects objects from images captured by the multiple cameras 120. In this embodiment, an example will be described in which a person is detected as an object.

[0012] 1, the control unit 110 includes an imaging control unit 111, a detection unit 112, a determination unit 113, an identification unit 114, a designation unit 115, and a communication control unit 116. A part of the functional configuration of the control unit 110 may be included in each of the multiple cameras 120. For example, the imaging control unit 111, the detection unit 112, and the determination unit 113 may be included in each of the multiple cameras 120 as functions of the camera 120 as an edge camera.

[0013] The shooting control unit 111 controls the first camera 121, the second camera 122, the third camera 123, and the fourth camera 124 to shoot a desired range and acquire the shot video. As an example, FIG. 2 shows the installation locations of the first camera 121, the second camera 122, the third camera 123, and the fourth camera 124. In the example shown in FIG. 2, the first camera 121, the second camera 122, the third camera 123, and the fourth camera 124 are arranged to shoot an intersection. The shooting spaces of the first camera 121, the second camera 122, the third camera 123, and the fourth camera 124 at least partially overlap with each other. In FIG. 2, the hatched area represents an intersection, a crosswalk, etc., and is an example of an area where people, etc., are monitored using multiple cameras 120. The hatched area may be an area where all of the shooting spaces of first camera 121, second camera 122, third camera 123, and fourth camera 124 overlap, as long as the shooting spaces of at least two cameras overlap. Furthermore, the shooting directions of first camera 121, second camera 122, third camera 123, and fourth camera 124 are different from each other. Note that in object detection system 100, the shooting directions of the multiple cameras may all be different, or some of them may be the same.

[0014] The detection unit 112 detects objects from images captured by the multiple cameras 120. Specifically, the detection unit 112 detects objects by performing object recognition processing on the images captured by the multiple cameras 120. In this embodiment, the detection unit 112 detects people as objects. In the following description, the detection unit 112 will be described as detecting people, and therefore the term "object" can be interchanged with "person."

[0015] The determination unit 113 determines whether or not the detected object is a monitoring target based on the state of the object detected by the detection unit 112. In this embodiment, a person is detected as the object, and therefore the determination unit 113 determines whether or not the person is a monitoring target based on the state of the person. Here, the state of a person that the determination unit 113 determines to be a monitoring target is, for example, a state in which the person is not paying attention to their surroundings. Furthermore, the state in which the person is not paying attention to their surroundings is, for example, a state in which the person is moving while operating a mobile terminal such as a smartphone, a state in which the person is running, or the like, in which the person is continuing to perform an action that prevents the person from checking their surroundings.

[0016] Specifically, the determination unit 113 detects the state of a person detected by the detection unit 112 by performing predetermined image processing on the area occupied by the person in the video captured by the camera 120. The determination unit 113 detects the configuration of the person from the area occupied by the person detected by the detection unit 112, and determines that the person is moving while operating a mobile device such as a smartphone based on the positional relationship of the configuration, for example, if the person's face is continuously facing downward, or if the person's face is continuously facing downward and the person's hands are in the direction the person's face is facing. The determination unit 113 also detects the person's movement speed from the fluctuation in position of the area occupied by the person detected by the detection unit 112 for each captured frame, and determines that the person is running if the movement speed is above a predetermined speed. Then, the determination unit 113 determines that the detected person is in a state where the person is paying less attention to their surroundings, and determines that the person is a monitoring target.

[0017] The identification unit 114 identifies the position of one of the multiple cameras 120 in the shooting space of the other cameras 120 based on the position of an object detected from a video captured by the one of the multiple cameras 120 in the shooting space of the one camera 120. Specifically, the identification unit 114 previously associates the spatial coordinates of the shooting spaces of the multiple cameras 120 with each other. More specifically, the identification unit 114 previously associates the spatial coordinates (also referred to as World coordinates) of the shooting spaces of the multiple cameras 120 with each other. Furthermore, the identification unit 114 previously associates the spatial coordinates (three-dimensional coordinates) of the shooting space of each camera 120 with the camera coordinates (three-dimensional coordinates) of the camera 120. Note that the image coordinates (two-dimensional coordinates) (x, y) in the video captured by the camera 120 correspond to the camera coordinates (x, y). As a result, the identification unit 114 can identify the spatial coordinates (position) of the object in the shooting space of the camera 120, based on the image coordinates of the object in the video captured by the camera 120. Then, the identification unit 114 can identify the spatial coordinates (position) of the object in the shooting space of the other camera 120, based on the identified spatial coordinates (position) of the object in the shooting space of the other camera 120. Furthermore, the identification unit 114 can identify the image coordinates of the object in the video captured by the other camera 120, based on the identified spatial coordinates (position) of the object in the shooting space of the other camera 120.

[0018] When an object detected from a video captured by one camera 120 is determined to be a monitoring target, and an object detected from a video captured by another camera 120 at a position in the shooting space of the other camera 120 that corresponds to the position in the shooting space of the first camera 120 of the object determined to be a monitoring target is not determined to be a monitoring target, the designation unit 115 designates the object detected from the video captured by the other camera 120 as a monitoring target. For example, when an object detected from a video captured by a first camera 121 is determined to be a monitoring target, the designation unit 115 determines whether an object is detected from the video captured by the second camera 122 at a position in the shooting space of the second camera 122 that corresponds to the position of the object in the shooting space of the first camera 121. In other words, the designation unit 115 determines whether or not there is an object detected from the video captured by the second camera 122 that is located at the same position in the shooting space as the object detected from the video captured by the first camera 121 and determined to be a monitoring target.

[0019] When an object detected from a video captured by second camera 122 has not been determined to be a monitoring target at a position in the shooting space of second camera 122 that corresponds to the position of the object in the shooting space of first camera 121, designation unit 115 designates the object detected from the video captured by second camera 122 as a monitoring target. In other words, even if the object detected from the video captured by second camera 122 has not been determined to be a monitoring target by determination unit 113, designation unit 115 designates the object detected from the video captured by second camera 122 as a monitoring target when the position in the shooting space is the same and the object detected from the video captured by first camera 121 has been determined to be a monitoring target by determination unit 113.

[0020] Communication control unit 116 controls communication unit 130 to transmit information regarding the status of a person determined to be a monitoring target to a vehicle (not shown) traveling through the intersection. Vehicles traveling through the intersection include vehicles traveling in the direction of entering the intersection monitored by object detection system 100 and vehicles traveling within the intersection. Specifically, when an object determined as a monitoring target by determination unit 113 or an object designated as a monitoring target by designation unit 115 is present in an image captured by any camera at the intersection monitored by object detection system 100, information regarding the status of the person determined to be a monitoring target is transmitted to the vehicle before it enters the intersection, for the period until it passes through the intersection.

[0021] The information about the state of the person transmitted by the communication control unit 116 may include information about the position of the person. Furthermore, the communication control unit 116 may control the communication unit 130 to transmit information about a warning to a mobile terminal carried by a person passing through the intersection.

[0022] The communication unit 130 is controlled by the communication control unit 116 and performs communication between the object detection system 100 and an external device. Specifically, the communication unit 130 transmits information about the status of a person determined to be a monitoring target to a vehicle (not shown) traveling through the intersection. The communication unit 130 may also transmit information about a warning to a mobile terminal carried by a person passing through the intersection.

[0023] Next, an object detection method performed by the object detection system 100 according to the first embodiment will be described with reference to Figures 3 and 4. The object detection method shown in Figures 3 and 4 is constantly executed by the object detection system 100.

[0024] First, determination unit 113 determines whether detection unit 112 has detected a person from the video captured by multiple cameras 120 (step S101). For convenience, in Fig. 3, it is described as determining whether a person has been detected from the video captured by first camera 121, and any camera among multiple cameras 120 included in object detection system 100 that has detected a person is referred to as first camera 121.

[0025] In step S101, if the determination unit 113 determines that the detection unit 112 has not detected a person from the images captured by the multiple cameras 120 (step S101; No), the process returns to step S101.

[0026] In step S101, if the determination unit 113 determines that the detection unit 112 has detected a person from the video captured by the multiple cameras 120 (step S101; Yes), the determination unit 113 determines whether or not the person is a monitoring target (step S102). In the following description, for example, it is assumed that a person is detected from the video captured by the first camera 121.

[0027] In step S102, if determination unit 113 determines that the person detected from the video captured by first camera 121 is not a surveillance target (step S102; No), the process returns to step S101.

[0028] In the processing of steps S101 and S102, tracking continues even for a person who was determined not to be a monitoring target in step S102, and the judgment unit 113 detects that a person who was determined not to be a monitoring target has become a monitoring target.

[0029] In step S102, if determination unit 113 determines that the person detected from the video captured by first camera 121 is a monitoring target (step S102; Yes), detection unit 112 continues to detect the person who is a monitoring target from the video sequentially captured by first camera 121 (step S103). That is, detection unit 112 tracks the person in the video sequentially captured by first camera 121.

[0030] Next, the identification unit 114 identifies the position of the person who has been detected from the video captured by the first camera 121 and determined to be a monitoring target in the shooting space of the other camera 120 (step S104). In the following description, the second camera 122 will be taken as an example of the other camera 120.

[0031] Next, the designation unit 115 determines whether the person detected from the video captured by the second camera 122 at the position in the shooting space of the second camera 122 identified in step S104 of the person determined to be a monitoring target in step S102 has already been determined to be a monitoring target (step S105).

[0032] In step S105, if the designation unit 115 determines that the person detected from the video captured by the second camera 122 has already been determined to be a monitoring target (step S105; Yes), the process proceeds to step S107.

[0033] In step S102, if in step S105 the designation unit 115 determines that the person detected from the video captured by the second camera 122 has not already been determined to be a monitoring target (step S105; No), the designation unit 115 designates the person as a monitoring target (step S106).

[0034] Next, detection unit 112 continues to detect the person who is the monitoring target from the images sequentially captured by second camera 122 (step S107). That is, detection unit 112 tracks the person in the images sequentially captured by second camera 122.

[0035] In step S106, if the designation unit 115 is tracking a person designated as a monitoring target in step S107, and the judgment unit 113 determines that the person is a monitoring target based on the image from the second camera 122, tracking continues in step S107.

[0036] Next, detection unit 112 determines whether or not to end tracking of the person being monitored from the images sequentially captured by multiple cameras 120, in this example, first camera 121 and second camera 122 (step S108). Specifically, detection unit 112 determines to end tracking of the person being monitored when the person being monitored is no longer detected from the images sequentially captured by first camera 121 and second camera 122.

[0037] In step S108, if the detection unit 112 determines not to end tracking of the person who is the monitoring target (step S108; No), the process of step S108 is repeated.

[0038] In step S108, if the detection unit 112 determines that tracking of the person who is the monitoring target is to be terminated (step S108; Yes), the process returns to step S101.

[0039] According to the object detection system 100 and object detection method according to the first embodiment described above, if a person who has been detected in the video captured by the first camera 121 and determined to be a monitoring target is detected in the video captured by the second camera 122 but is not determined to be a monitoring target, the person detected in the video captured by the second camera 122 is designated as a monitoring target by the designation unit 115. When the shooting directions of the multiple cameras 120 are different, there are cases where the same pedestrian is determined to be a monitoring target in the video captured by one camera but not in the video captured by the other camera.

[0040] For example, when the shooting directions of the multiple cameras 120 are different, the orientation of the cameras 120 of the person relative to the light source is different, and therefore the determination unit 113 may not be able to appropriately detect the state of the person. Specifically, for example, when the first camera 121 captures the front of the person and the second camera 122 captures the back of the person, even if the determination unit 113 can detect the state of the person operating the smartphone from the video captured by the first camera 121, it may not be able to detect the state of the person operating the smartphone from the video captured by the second camera 122. Furthermore, for example, if the first camera 121 captures the right side of the person and the second camera 122 captures the left side of the person, the judgment unit 113 may be able to detect that the person is operating a smartphone from the image captured by the first camera 121, but may not be able to detect that the person is operating a smartphone from the image captured by the second camera 122, for example, if the person is photographed in a dark image due to backlighting or in a blown-out image due to direct sunlight.

[0041] On the other hand, according to the object detection system 100 and the object detection method according to the first embodiment, the designation unit 115 designates as a monitoring target a person who is detected from a video captured by any one of the multiple cameras 120 (for example, the first camera 121) and determined to be a monitoring target, and who is detected from a video captured by another of the multiple cameras 120 (for example, the second camera 122). Therefore, even if the shooting directions of the multiple cameras 120 are different, it is possible to detect as a monitoring target a person who is detected from a video captured by each camera 120 and who should be a monitoring target. Therefore, it is possible to provide the object detection system 100 and the object detection method that can appropriately monitor a monitoring target with the multiple cameras 120.

[0042] Embodiment 2 Next, an object detection system 100A according to a second embodiment of the present invention will be described. Fig. 5 shows an example of the configuration of the object detection system 100A according to the second embodiment. The object detection system 100A according to the second embodiment differs from the object detection system 100 according to the first embodiment in that it further includes an image quality adjustment unit 117. Therefore, among the configuration of the object detection system 100A according to the second embodiment, the same configuration as that of the object detection system 100 according to the first embodiment will be assigned the same reference numerals, and their description will be omitted.

[0043] Image quality adjustment unit 117 adjusts the image quality of the images captured by multiple cameras 120. Furthermore, image quality adjustment unit 117 adjusts the image quality of an area in the images captured by the other cameras 120 that corresponds to an object detected in the image captured by one camera 120 and determined to be a monitoring target, thereby improving the object detection accuracy. Specifically, when it is determined that an object detected in the image captured by one camera 120 is a monitoring target, image quality adjustment unit 117 adjusts the image quality of an area in the images captured by the other cameras 120 that corresponds to the object, thereby improving the object detection accuracy. For example, when it is determined that an object detected in the image captured by first camera 121 is a monitoring target, image quality adjustment unit 117 adjusts the image quality of an area occupied by the object in the image captured by second camera 122.

[0044] The image quality adjustment performed by the image quality adjustment unit 117 to improve the object detection accuracy can be a combination of any image processing such as contrast adjustment processing, gain adjustment processing, and edge enhancement processing.

[0045] Next, an object detection method performed by object detection system 100A according to embodiment 2 will be described with reference to Fig. 6. As shown in Fig. 6, the object detection method according to embodiment 2 differs from the object detection method according to embodiment 1 in that the processing of step S201 is included between steps S106 and S107. Therefore, steps in the object detection method according to embodiment 2 that are the same as those in the object detection method according to embodiment 1 are assigned the same step numbers, and descriptions thereof will be omitted.

[0046] After step S106, image quality adjustment unit 117 performs image quality adjustment on the range of the person who is the monitoring target in the image captured by second camera 122 (step S201).

[0047] The process of step S201 may determine what kind of image quality adjustment to perform depending on the state of the image of the range of the person who is the monitoring target in the video captured by second camera 122. For example, if the state of the image of the range of the person who is the monitoring target in the video captured by second camera 122 is an image with low brightness overall, an adjustment to increase the gain or contrast overall is performed. Furthermore, if the state of the image of the range of the person who is the monitoring target in the video captured by second camera 122 is an image with high brightness overall and has blown-out highlights, an adjustment to decrease the gain overall is performed. Furthermore, the process of step S201 may adjust the gain or the like so that the state of the image of the range of the person who is the monitoring target has characteristics similar to the image of the range of the person who is determined to be the monitoring target in the video captured by first camera 121. Furthermore, in addition to the above image quality adjustment, edge enhancement processing or the like may be combined.

[0048] According to the object detection system 100A and the object detection method of the second embodiment described above, similarly to the object detection system 100 and the object detection method of the first embodiment, even when the shooting directions of the multiple cameras 120 are different, it is possible to detect as a monitoring target a person who is detected from the video captured by each camera 120 and who should be monitored. Furthermore, image quality adjustment is performed on the range of the person designated as the monitoring target in the video captured by the second camera 122, thereby improving the detection accuracy of the person. This allows the determination unit 113 to accurately determine that a person who should be monitored is a monitoring target from the video sequentially captured by the second camera 122. Therefore, the determination unit 113 can determine from the beginning that a person designated as a monitoring target in the video captured by the second camera 122 by the designation unit 115 is a monitoring target in the video subsequently sequentially captured by the second camera 122.

[0049] Embodiment 3 Next, an object detection system 100B according to a third embodiment of the present invention will be described. Fig. 7 shows an example of the configuration of object detection system 100B according to the third embodiment. Object detection system 100B according to the third embodiment differs from object detection system 100 according to the first embodiment in that it further includes an angle-of-view changer 118. Therefore, among the configuration of object detection system 100B according to the third embodiment, the same configuration as that of object detection system 100 according to the first embodiment will be denoted by the same reference numerals, and description thereof will be omitted.

[0050] The angle of view changing unit 118 changes the angles of view of the multiple cameras 120. The angle of view changing unit 118 changes the angles of view of the other cameras 120 so as to enlarge and capture a range in the video captured by the other cameras 120 that corresponds to an object detected in the video captured by one camera 120 and determined to be a monitoring target. Specifically, when it is determined that an object detected in the video captured by one camera 120 is a monitoring target, the angle of view changing unit 118 changes the angle of view of the other cameras 120 so as to enlarge and capture a range in the video captured by the other cameras 120 that corresponds to the object. For example, when it is determined that an object detected in the video captured by the first camera 121 is a monitoring target, the angle of view changing unit 118 changes the angle of view of the second camera 122 so as to enlarge and capture a range occupied by the object in the video captured by the second camera 122.

[0051] The change in the angle of view of the other camera 120 by the angle of view changer 118 is a change to capture an image so as to enlarge the range occupied by a detected object in the captured video, specifically, zoom capture. If such a change in the angle of view is optical zoom capture, the camera 120 is provided with an optical zoom function, and the zoom operation is controlled by the capture control unit 111. If the change in the angle of view is a process of cutting out and enlarging the range occupied by a detected object in the captured video, processing is performed by the angle of view changer 118. In this case, interpolation processing such as super-resolution processing may be performed on the cut-out range.

[0052] Next, an object detection method performed by object detection system 100B according to the third embodiment will be described with reference to Fig. 8. As shown in Fig. 8, the object detection method according to the third embodiment differs from the object detection method according to the first embodiment in that the processing of step S301 is included between steps S106 and S107. Therefore, steps in the object detection method according to the third embodiment that are the same as those in the object detection method according to the first embodiment are assigned the same step numbers, and descriptions thereof will be omitted.

[0053] After step S106, the angle of view changer 118 changes the angle of view of the second camera 122 so as to enlarge the range of the person who is the monitoring target in the image of the second camera 122 (step S301).

[0054] The process of step S301 may determine whether or not enlargement is necessary based on the image size of the range of the person designated as the monitoring target in the image captured by second camera 122. For example, this applies when a detected person is detected in a position close to first camera 121 and is determined to be the monitoring target from the image captured by first camera 121, and the range of the person designated as the monitoring target is small in the image captured by second camera 122 because the distance to the person designated as the monitoring target is far. Specifically, this is effective when the image size of the range of the person designated as the monitoring target in the image captured by second camera 122 by designation unit 115 is two-thirds or less in the vertical or horizontal direction of the image size of the range of the person designated as the monitoring target in first camera 121. Furthermore, even if the image size of the range of a person determined to be a monitoring target by the first camera 121 is the same as the image size of the range of a person determined to be a monitoring target in the image of the second camera 122 by the designation unit 115, by enlarging and photographing the range of the person determined to be a monitoring target, the judgment unit 113 may be able to determine that a person who could not be determined to be a monitoring target based on the image of the second camera 122 is a monitoring target.

[0055] The object detection system 100B and object detection method according to the third embodiment described above can achieve the same effects as the object detection system 100 and object detection method according to the first embodiment. Furthermore, since the angle of view of the second camera 122 is changed so as to expand the range of a person designated as a monitoring target in the video captured by the second camera 122, the accuracy of detecting the person can be improved. This enables the determination unit 113 to accurately determine that a person who should be monitored is a monitoring target from the video sequentially captured by the second camera 122. Therefore, the determination unit 113 can determine, from the beginning, that a person designated as a monitoring target in the video captured by the second camera 122 by the designation unit 115 is a monitoring target in the video subsequently sequentially captured by the second camera 122.

[0056] In the above-described embodiment, the present invention has been described as being configured as hardware, but the present invention is not limited to this. For example, the present invention can be realized by having a CPU execute a computer program to perform the above-described processing procedures shown in Figures 3, 4, 6, 8, etc.

[0057] The above-described 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-R / Ws, and semiconductor memories (e.g., mask ROMs, PROMs (Programmable ROMs), EPROMs (Erasable PROMs), flash ROMs, and RAMs (Random Access Memory)). The program can 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 transitory computer-readable media can supply the program to a computer via a wired communication path such as an electric wire or optical fiber, or via a wireless communication path.

[0058] The present invention is not limited to the above-described embodiment, and can be modified as appropriate without departing from the spirit of the present invention. For example, image quality adjustment unit 117 may adjust the image quality of the range of the person who is the monitoring target in the video captured by first camera 121 as well as second camera 122. Furthermore, angle of view change unit 118 may change the angle of view of first camera 121 so as to enlarge the range of the person who is the monitoring target in the video captured by first camera 121 as well as second camera 122. [Explanation of symbols]

[0059] 100, 100A, 100B Object Detection System 111 Shooting control unit 112 Detection unit 113 Judgment Department 114 Specific section 115 Designated part 116 Communication control unit 117 Image quality adjustment section 118 Angle of view change section 121 Camera 1 122 Second Camera 123 Third Camera 124 4th Camera 130 Communications Department

Claims

1. An object detection system comprising a plurality of cameras, wherein at least a portion of the imaging spaces of the plurality of cameras overlap with each other, a detection unit that detects an object from images captured by the plurality of cameras; a determination unit that determines whether or not the object is a monitoring target based on the state of the object detected by the detection unit; an identification unit that associates spatial coordinates of the respective shooting spaces of the plurality of cameras with each other, and identifies a position of an object detected from a video captured by one of the plurality of cameras in the shooting space of the other cameras based on a position of the object in the shooting space of the one camera; a designation unit that designates an object detected from the video captured by the one camera as a monitoring target when the object detected from the video captured by the other camera is determined to be a monitoring target, and when the object detected from the video captured by the other camera is not determined to be a monitoring target at a position in the shooting space of the other camera that corresponds to the position in the shooting space of the one camera of the object determined to be a monitoring target; Equipped with Object detection system.

2. further comprising an image quality adjustment unit that adjusts image quality of the images captured by the plurality of cameras; the image quality adjustment unit performs image quality adjustment to improve object detection accuracy for a range in the image captured by the other camera that corresponds to an object detected from the image captured by the one camera and determined to be a monitoring target; The object detection system of claim 1 .

3. a view angle changing unit that changes the view angles of the plurality of cameras; the angle of view change unit changes the angle of view of the other camera so as to enlarge and capture an area in the video captured by the other camera that corresponds to an object that has been detected from the video captured by the one camera and that has been determined to be a monitoring target; The object detection system of claim 1 .

4. the detection unit detects a person as the object, the determination unit determines that the person is a surveillance target when the state of the person as the state of the object is a state in which the person is not paying attention to their surroundings; The object detection system according to any one of claims 1 to 3.

5. The computer a detection step of detecting an object from images captured by a plurality of cameras whose capture spaces at least partially overlap each other; a determination step of determining whether or not the object is a monitoring target based on the state of the object detected in the detection step; an identifying step of associating spatial coordinates of the respective shooting spaces of the plurality of cameras with each other, and identifying a position of an object detected from an image captured by one of the plurality of cameras in the shooting space of the one camera based on the position of the object in the shooting space of the one camera; a designation step of designating the object detected from the video captured by the one camera as a monitoring target when the object detected from the video captured by the other camera is determined to be a monitoring target, and when the object detected from the video captured by the other camera is not determined to be a monitoring target at a position in the shooting space of the other camera corresponding to the position in the shooting space of the one camera of the object determined to be a monitoring target; ,Object detection method.

Images