Information processing device, information processing method, and inspection system

Abstract

An information processing device according to one embodiment of the present technology is provided with an image generation unit, a detection unit, and a calculation unit. The image generation unit generates the entire image on the basis of event data acquired from an event-based vision sensor (EVS). The detection unit detects an inspection target on the basis of the entire image. The calculation unit calculates position information about the inspection object on the basis of the detection result from the detection unit. Thereby, it is possible to inspect an object, which is an inspection object in an infrastructure inspection or the like, with high accuracy.

Description

Information processing device, information processing method, and inspection system

[0001] The present technology relates to an information processing device, an information processing method, and an inspection system that can be applied to the inspection of an object or the like.

[0002] Patent Document 1 describes an imaging system that includes an overall imaging camera mounted on a mobile body that captures an entire range of a predetermined angle of view to generate an overall image, and multiple partial imaging cameras mounted on the mobile body that capture multiple ranges that divide the entire range of the predetermined angle of view to generate partial images, and that generates data that associates the overall image and partial images captured at the same time.This makes it possible to inspect objects scattered over a wide area using the images of those objects (see, for example, paragraphs

[0026] to

[0036] and Figures 1 to 3 of the specification of Patent Document 1).

[0003] Japanese Patent Application Laid-Open No. 2019-121821

[0004] There is a demand for an information processing device, an information processing method, and an inspection system that are capable of inspecting objects that are the targets of inspection in infrastructure inspections and the like with high accuracy.

[0005] In view of the above circumstances, an object of the present technology is to provide an information processing device, an information processing method, and an inspection system that are capable of inspecting an object with high accuracy.

[0006] To achieve the above object, an information processing device according to an embodiment of the present technology includes an image generation unit, a detection unit, and a calculation unit. The image generation unit generates an overall image based on event data acquired from an EVS (Event-based Vision Sensor). The detection unit detects an inspection object based on the overall image. The calculation unit calculates position information of the inspection object based on a detection result by the detection unit.

[0007] This information processing device generates an overall image based on event data acquired from the EVS, and calculates the position information of the detected inspection object from the overall image, thereby enabling high-precision inspection of the object.

[0008] The information processing device may further include a control unit that controls, based on position information of the inspection object, a partial camera that can capture an image of a region corresponding to the position information.

[0009] The control unit may control the partial cameras to capture an image of a region that includes an outer shape of the inspection target.

[0010] When the inspection object is detected, the control unit may select a part of the partial cameras that can capture an image of the area, and cause the partial cameras to capture an image of the area.

[0011] The information processing device may further include an inspection unit that inspects the inspection object for damage based on one or more partial images including the inspection object captured by the partial camera.

[0012] When the inspection object is detected, the control unit may generate a timestamp indicating the time when the inspection object is detected.

[0013] The timestamp may be assigned to a group of partial images captured by all of the partial cameras.

[0014] The information processing device may further include a storage unit that stores inspection information related to inspection of damage of the inspection object, in which the inspection information may include the partial image group to which the timestamp is assigned and position information of the inspection object.

[0015] The inspection information may include embedded data (ED) in the EVS or ED in the partial camera.

[0016] When the timestamp is assigned to the partial image group, the storage unit may store the partial image group to which the timestamp is assigned and the ED in which position information of the inspection object is output.

[0017] The detection unit may detect the inspection target using a trained model for detecting the inspection target, in which case the trained model may be made of a neural network.

[0018] The detection unit may detect whether or not the object to be inspected is damaged using the trained model to which learning data consisting of images relating to the damage to the object to be inspected is input as a parameter.

[0019] The detection unit may detect whether or not the inspection object is damaged based on the entire image.

[0020] According to an embodiment of the present technology, an information processing method is executed by a computer system, and includes generating an overall image based on event data acquired from an event-based vision sensor (EVS), detecting an inspection object based on the overall image, and calculating position information of the inspection object based on a detection result by the detection unit.

[0021] To achieve the above object, an inspection system according to one embodiment of the present technology includes an EVS (Event-based Vision Sensor) and an information processing device. The information processing device includes an image generation unit, a detection unit, and a calculation unit. The image generation unit generates an overall image based on event data acquired from the EVS (Event-based Vision Sensor). The detection unit detects an inspection object based on the overall image. The calculation unit calculates position information of the inspection object based on a detection result by the detection unit.

[0022] The information processing device may further include a control unit that controls a partial camera that can capture an image of an area corresponding to the position information of the inspection object based on the position information of the inspection object. In this case, the partial camera may be a plurality of cameras.

[0023] Fig. 1 is a diagram schematically showing an inspection system; Fig. 2 is a diagram showing a flowchart of an inspection for the presence or absence of damage to an inspection object; Fig. 3 is a diagram schematically showing an entire image and a partial image; Fig. 4 is a diagram schematically showing a partial image; Fig. 4 is a diagram schematically showing an inspection system; Fig. 5 is a diagram showing another flowchart of an inspection for the presence or absence of damage to an inspection object;

[0024] Hereinafter, embodiments of the present technology will be described with reference to the drawings.

[0025] First Embodiment FIG. 1 is a diagram schematically illustrating an inspection system 100 according to a first embodiment of the present technology.

[0026] 1, the inspection system 100 includes a mobile object 1 equipped with an overall image capturing camera 5, a partial image capturing camera 6, and an inspection device 20, and an information processing device 10. The inspection system 100 also captures an image of the inspection target to inspect whether or not the inspection target is damaged.

[0027] Typically, inspection targets include various facilities and equipment related to infrastructure, such as tracks (railroad tracks), roads, tunnels, runways, bridges, houses, dams, sewers, utility poles, electric wires, signs, etc. Inspection targets also include locations (candidates) in the above infrastructure where damage (cracks, corrosion, water leaks, etc.) may have occurred.

[0028] Note that the inspection targets may also include objects other than infrastructure, such as trees or fallen trees in contact with infrastructure, bird droppings or nests, falling rocks, and other objects that may obstruct or cause obstructions to infrastructure.

[0029] It is desirable that the mobile body 1 be able to move along the shape of the object to be inspected. For example, various mobile bodies 1 may be used depending on the type of object to be inspected, such as a train for inspecting tracks, a car for inspecting roads and tunnels, and a drone for inspecting utility poles and bridges.

[0030] The overall imaging camera 5 captures an image of a wide area including the inspection target. In this embodiment, the overall imaging camera 5 is an event-based vision sensor (EVS). The EVS is a sensor that asynchronously detects changes in pixel luminance and outputs event data. Here, the event data is data including the coordinates (X coordinate, Y coordinate) of the pixel where the luminance change occurred, the time when the change was detected, and the polarity of the luminance change.

[0031] A plurality of partial imaging cameras 6 are arranged, and capture images of any of the ranges obtained by dividing the area captured by the overall imaging camera 5 into the same number of parts as the number of cameras arranged. For example, in Fig. 3B, the overall area 50 captured by the overall imaging camera 5 is illustrated as being divided into 30 divided areas (A1 to A30). In other words, 30 partial imaging cameras 6 are arranged. Hereinafter, an image captured by a partial imaging camera 6 will be referred to as a partial image, and a plurality of partial images (e.g., A1 and A2) will be referred to as a group of partial images.

[0032] Furthermore, the overall imaging camera 5 and the partial imaging camera 6 are disposed at any position depending on the position of the inspection target. For example, if the inspection target is located ahead of the mobile body 1, they may be disposed on the front of the mobile body 1. Of course, the placement of the overall imaging camera 5 and the partial imaging camera 6 is not limited to one direction, and for example, when inspecting a tunnel by car, they may be disposed in each direction, such as the front for capturing images in the direction of travel, the left and right sides for capturing images perpendicular to the direction of travel, or the top for inspecting the upper part of the tunnel.

[0033] The information processing device 10 includes a processor 15 having an overall image generation unit 11, a detection unit 12, a partial camera control unit 13, and a partial image acquisition unit 14, a position measurement unit 16, and a storage unit 17. In this embodiment, the information processing device 10 is mounted on a moving object 1.

[0034] The processor 15 controls the operation of each part of the information processing device 10 by executing processes according to the programs.

[0035] The overall image generating unit 11 generates an image based on the event data output from the EVS. Hereinafter, the image generated by the overall image generating unit 11 will be referred to as an overall image.

[0036] The detection unit 12 detects the inspection object within the entire image. In this embodiment, the detection unit 12 detects the inspection object using an object detection neural network. In this embodiment, in an initial setup stage, various images related to the inspection object are input to the learning function of the object detection neural network to perform learning, and learning result data obtained as a result of the learning is incorporated as parameters.

[0037] This makes it possible to automatically and quickly determine whether candidate regions of various sizes resemble pre-trained objects (inspection targets).Furthermore, it is possible to improve the accuracy of judgment compared to applying a neural network to pre-divided regions of fixed size.

[0038] The method used for the object detection neural network is not limited, and R-CNN, Faster R-CNN, Single Shot MultiBox Detector, etc. may be used.

[0039] The partial camera control unit 13 calculates position information of the inspection target based on the detection result by the detection unit 12. The partial camera control unit 13 also selects a partial imaging camera 6 according to the position information of the inspection target. A specific example will be described later with reference to FIGS. 2 and 3.

[0040] The partial image capturing camera 6 selected by the partial camera control unit 13 captures an image of the corresponding divided area. The partial image acquisition unit 14 acquires the partial image captured by the partial image capturing camera 6.

[0041] The position measurement unit 16 measures the position of the mobile body 1. For example, the position measurement unit 16 uses a GNSS (Global Navigation Satellite System) unit and measures the position of the mobile body 1 on which the information processing device 10 is mounted by receiving navigation signals transmitted from artificial satellites.

[0042] The storage unit 17 stores the partial images acquired by the partial image acquisition unit 14 in association with the position information of the mobile object 1. The storage unit 17 may also store the position information when the inspection target is imaged and the partial images as a DB. For example, the date and time, position, partial image, inspection result, cause of damage, etc. may be associated and stored.

[0043] The stored partial image (image showing the inspection object) is output to the inspection device 20. The inspection device 20 inspects the inspection object for the presence or absence of a fault (damage). For example, the inspection device 20 has a speed detection unit that detects speed information (including speed and acceleration) of the mobile object 1, a recording unit that records the type of damage and position information, etc. Note that the inspection device 20 does not need to be mounted on the mobile object 1 or the information processing device 10, but may be arranged externally.

[0044] Fig. 2 is a diagram showing a flowchart of an inspection for the presence or absence of damage to an inspection object. Fig. 3 is a diagram showing a schematic view of an entire image and a partial image. Fig. 3A is a diagram showing a schematic view of an entire image. Fig. 3B is a diagram showing a schematic view of a partial image.

[0045] As shown in FIG. 2, an image is taken by the overall image capturing camera 5 (EVS) (step 101).

[0046] The whole image generating unit 11 generates a whole image based on the event data output from the EVS (step 102).

[0047] The detection unit 12 detects the inspection objects from the generated overall image (step 103). For example, as shown in Fig. 3A, inspection objects 30a, 30b, 30c, and 30d are detected in the overall image 50. In this embodiment, the object detection neural network of the detection unit 12 obtains the position coordinates of a rectangle (see dotted line) surrounding the inspection objects.

[0048] The partial camera control unit 13 calculates position information of the inspection object based on the detection result by the detection unit 12 (step 104). The partial camera control unit 13 also selects a partial imaging camera 6 corresponding to the coordinates of the inspection object (step 105). For example, as shown in FIG. 3B , the inspection object 30a is located at coordinates A8, A9, A14, A15, A20, A21, A26, and A27. That is, the partial camera control unit 13 selects a corresponding partial imaging camera 6 so as to image the above-mentioned divided area that includes the outline of the inspection object 30a.

[0049] The selected partial-image capturing camera 6 captures a partial image of the corresponding divided area (step 106). The partial-image acquiring unit 14 outputs the captured partial image to the storage unit 17. The storage unit 17 then stores the partial images (step 107). That is, the storage unit 17 stores a group of partial images A8, A9, A14, A15, A20, A21, A26, and A27 captured from the inspection object 30a, a group of partial images A9 and A15 captured from the inspection object 30b, a group of partial images A10 and A16 captured from the inspection object 30c, and a group of partial images A11, A12, A17, A18, A23, A24, A29, and A30 captured from the inspection object 30d.

[0050] The stored partial image is output to the inspection device 20, which detects whether or not there is a fault in the object to be inspected (step 108).

[0051] As described above, the information processing device 10 according to this embodiment generates an overall image based on event data acquired from the overall image capture camera 5 (EVS), and calculates position information of the inspection target detected from the overall image, thereby enabling high-precision inspection of the object.

[0052] Conventional imaging systems for inspecting infrastructure from mobile vehicles have been equipped with a camera for capturing an overall image of inspection objects scattered over a wide area, and multiple cameras for capturing partial images of each of multiple areas divided by a predetermined angle of view. This method consumes a lot of power because cameras that do not capture the inspection object are also in operation. Furthermore, the amount of data increases because all image data is output to an information processing device. Furthermore, because pixels that do not capture the inspection object are also read, the frame rate cannot be increased, resulting in low inspection accuracy and efficiency. Furthermore, inspection efficiency is low because inspectors have to silently identify the inspection object from the overall image.

[0053] In this technology, an EVS is used for the camera for capturing the entire image. The EVS is a sensor that reacts only to pixels that have a change in brightness and outputs the data. The presence of an object to be inspected is determined from the image captured by the EVS, and only the partial imaging camera that captures the object to be inspected is activated to capture the object. Furthermore, the imaging angle of the partial imaging camera is limited to the angle that includes the object to be inspected.

[0054] As a result, using an EVS for capturing the entire image can reduce power consumption compared to capturing the entire image. Also, by operating only the partial image capturing camera that captures the object under inspection, power consumption can be reduced, and the amount of image data output can be reduced, so the amount of data processing can also be reduced. Furthermore, when capturing an image of an object under inspection, capturing only the area where the object under inspection is located can increase the frame rate, improving inspection efficiency and accuracy. Furthermore, since the object under inspection is detected within the imaging system, no inspector is required, which improves inspection efficiency.

[0055] Second Embodiment An information processing device 10 according to a second embodiment of the present technology will be described. In the following description, descriptions of parts having the same configurations and functions as those of the information processing device 10 described in the above embodiment will be omitted or simplified.

[0056] In the above embodiment, only the partial image capturing camera 6 selected by the partial camera control unit 13 was controlled to capture images. However, this is not limiting, and the partial image capturing camera 6 may be kept operating at all times, and all partial images may be recorded in the storage unit 17.

[0057] In this case, the overall image captured by the overall imaging camera 5 is output to the detection unit, and the inspection object is detected from the overall image. The detection unit 12 outputs a recognition signal indicating that the inspection object has been detected to the partial camera control unit 13. When the recognition signal is input, the partial camera control unit 13 assigns a timestamp to the partial image captured by the partial imaging camera 6. That is, in the flowchart shown in FIG. 2, the timestamp is assigned after step 103. The image with the timestamp assigned is stored in the storage unit 17 together with its location information. This enables efficient infrastructure inspection.

[0058] In the second embodiment, a timestamp is added to a partial image when an inspection object is detected. However, the present invention is not limited to this. When a timestamp is added, position information may be output in Embedded Data together with the partial image.

[0059] Fig. 4 is a diagram showing a schematic representation of a partial image. In Fig. 4, the partial image capturing camera 6 is always in operation, so all of the divided areas are captured, resulting in an image that is identical to the entire image. Note that in Fig. 4, the frame lines for each divided area are omitted.

[0060] 4, when inspection object 30a, inspection object 30b, inspection object 30c, and inspection object 30d are detected, a time stamp is assigned to each of them. Embedded data in which the position information of the inspection objects is output is stored in the storage unit 17 together with the partial images to which the time stamp is assigned.

[0061] The embedded data includes various information related to the sensor (imaging device), such as the frame count value (time code), gain value (analog / digital), white balance, exposure time, HDR information, and temperature and power supply monitor information of the sensor.

[0062] If a timestamp is not added, the partial image is not output to the storage unit 17 in order to reduce the amount of data. If it is determined that the inspection object has a fault, the determination result as to whether or not there is a fault may be included in the Embedded Data, similar to the timestamp.

[0063] In the second embodiment, among the partial images (A1 to A30) to which time stamps have been added, partial images that do not include the inspection target may be deleted.

[0064] In the above embodiment, the presence or absence of damage to the inspection object is inspected from the partial image by the external inspection device 20. However, the present invention is not limited to this, and the presence or absence of damage to the inspection object may be inspected by an object neural network of the detection unit.

[0065] FIG. 5 is a diagram schematically illustrating an inspection system 200 according to the fourth embodiment.

[0066] 5 includes an overall image capturing camera 5 and an information processing device 210. The information processing device 210 includes a processor 15 having an overall image generating unit 11 and a detecting unit 12, a position measuring unit 16, and a storage unit 17.

[0067] In FIG. 5, in the initial setting stage of the information processing device 210, various images related to the inspection object are input to the learning function of the object detection neural network of the detection unit 12 to cause learning, and the learning result data obtained as a result of the learning is input as a parameter.

[0068] For example, learning is performed by inputting images of cracks, corrosion, breaks, deterioration, etc. of inspection objects such as roads, utility poles, and tunnels.

[0069] This makes it possible to detect a fault in the inspection target from the overall image acquired by the overall image capturing camera 5 (EVS), which means that the partial image capturing camera 6 and the partial camera control unit 13 can be omitted.

[0070] FIG. 6 is a diagram showing another flowchart of the inspection for the presence or absence of damage to the inspection object.

[0071] 6, an image is captured by the overall image capturing camera 5 (EVS) (step 201). An overall image is generated by the overall image generating unit 11 based on the event data output from the EVS (step 202).

[0072] The detection unit 12, which has the object detection neural network that has been trained as described above, detects a fault in the inspection object from the overall image (step 203). Furthermore, the position information and image data of the inspection object in which a fault has been detected are transmitted to the storage unit 17 (step 204). Note that the storage unit 17 may store not only the position information and image data but also the time (timestamp) at which the fault was detected, the type of fault, the cause of the fault, and so forth.

[0073] Other Embodiments The present technology is not limited to the above-described embodiments, and various other embodiments can be realized.

[0074] In the above embodiment, the inspection of the inspection object for damage was performed by outputting a group of partial images of the inspection object from the information processing device 10 to the inspection device 20. Without being limited to this, the information processing device 10 may have a display unit on which the user can visually check the inspection results of the inspection object, an operation unit that accepts various input operations by the user, etc. For example, the user may be able to start the inspection of the inspection object by operating the operation unit of the information processing device 10 mounted on the mobile object 1.

[0075] The configurations of the image generating unit, the detecting unit, the calculating unit, etc. described with reference to the drawings are merely one embodiment and can be arbitrarily modified without departing from the spirit of the present technology. In other words, any other configurations, algorithms, etc. for implementing the present technology may be adopted.

[0076] It should be noted that the effects described in this disclosure are merely examples and are not limiting, and other effects may also be present. The description of multiple effects above does not necessarily mean that these effects are exhibited simultaneously. It means that at least one of the effects described above can be obtained depending on the conditions, etc., and of course, effects not described in this disclosure may also be exhibited.

[0077] It is also possible to combine at least two of the characteristic features of each embodiment described above. In other words, the various characteristic features described in each embodiment may be combined in any manner without distinguishing between the embodiments.

[0078] The present technology can also be configured as follows. (1) An information processing device comprising: an image generation unit that generates an overall image based on event data acquired from an EVS (Event-based Vision Sensor); a detection unit that detects an inspection object based on the overall image; and a calculation unit that calculates position information of the inspection object based on a detection result by the detection unit. (2) The information processing device described in (1), further comprising: a control unit that controls partial cameras that can capture an image of an area corresponding to the position information of the inspection object based on the position information of the inspection object. (3) The information processing device described in (2), wherein there are multiple partial cameras, and the control unit controls the partial cameras to capture an image of an area that includes the outline of the inspection object. (4) The information processing device described in (3), wherein, when the inspection object is detected, the control unit selects some of the partial cameras that can capture an image of the area and causes them to capture an image of the area. (5) The information processing device according to (4), further comprising an inspection unit that inspects the inspection object for damage based on one or more partial images including the inspection object captured by the partial camera. (6) The information processing device according to (2), wherein the control unit generates a timestamp indicating the time when the inspection object is detected when the inspection object is detected. (7) The information processing device according to (6), wherein the timestamp is assigned to a group of partial images captured by all of the partial cameras. (8) The information processing device according to (7), further comprising a storage unit that stores inspection information related to the inspection of damage to the inspection object, wherein the inspection information includes the group of partial images to which the timestamp is assigned and position information of the inspection object. (9) The information processing device according to (8), wherein the inspection information includes ED (Embedded Data) in the EVS or ED in the partial camera.(10) The information processing device according to (9), wherein, when the timestamp is assigned to the partial image group, the storage unit stores the partial image group to which the timestamp is assigned and the ED in which position information of the inspection object is output. (11) The information processing device according to (1), wherein the detection unit detects the inspection object using a trained model for detecting the inspection object, and the trained model is made up of a neural network. (12) The information processing device according to (11), wherein the detection unit detects the presence or absence of damage to the inspection object using the trained model to which training data made up of images related to damage to the inspection object is input as a parameter. (13) The information processing device according to (12), wherein the detection unit detects the presence or absence of damage to the inspection object based on the entire image. (14) An information processing method in which a computer system executes the following: generating an overall image based on event data acquired from an EVS (Event-based Vision Sensor), detecting an inspection object based on the overall image, and calculating position information of the inspection object based on a detection result by the detection unit. (15) An inspection system comprising: an EVS (Event-based Vision Sensor), and an information processing device having: an image generation unit that generates an overall image based on event data acquired from the EVS, a detection unit that detects the inspection object based on the overall image, and a calculation unit that calculates position information of the inspection object based on a detection result by the detection unit. (16) The inspection system according to (15), in which the information processing device has a control unit that controls a partial camera that can capture an image of an area corresponding to the position information of the inspection object, based on the position information of the inspection object, and there are multiple partial cameras.

[0079] DESCRIPTION OF SYMBOLS 1... Mobile body 5... Whole image capturing camera 6... Partial image capturing camera 10... Information processing device 11... Whole image generating unit 12... Detection unit 13... Partial camera control unit 20... Inspection device 100... Inspection system

Claims

1. An information processing device comprising: an image generation unit that generates an overall image based on event data acquired from an EVS (Event-based Vision Sensor); a detection unit that detects an inspection object based on the overall image; and a calculation unit that calculates position information of the inspection object based on the detection result by the detection unit.

2. An information processing device according to claim 1, further comprising: a control unit for controlling a partial camera capable of capturing an image of an area corresponding to said position information based on said position information of the object to be inspected.

3. An information processing device according to claim 2, wherein the partial cameras are multiple, and the control unit controls the partial cameras to capture an image of an area that includes the outer shape of the inspection object.

4. An information processing device according to claim 3, wherein the control unit, when the inspection object is detected, selects a part of the partial cameras capable of capturing an image of the area, and causes the partial cameras to capture an image of the area.

5. An information processing device according to claim 4, further comprising an inspection unit that inspects the inspection object for damage based on one or more partial images including the inspection object captured by the partial camera.

6. An information processing device according to claim 2, wherein the control unit generates a timestamp indicating a time when the inspection object is detected.

7. An information processing device according to claim 6, wherein the time stamp is assigned to a group of partial images captured by all of the partial cameras.

8. An information processing device according to claim 7, further comprising a memory unit for storing inspection information relating to inspection of damage to the inspection object, the inspection information including the group of partial images to which the timestamp has been assigned, and position information of the inspection object.

9. An information processing device according to claim 8, wherein the inspection information includes ED (Embedded Data) in the EVS or ED in the partial camera.

10. An information processing device as described in claim 9, wherein when the timestamp is assigned to the partial image group, the memory unit stores the partial image group to which the timestamp has been assigned and the ED in which positional information of the inspection object is output.

11. An information processing device according to claim 1, wherein the detection unit detects the inspection object using a trained model for detecting the inspection object, and the trained model is composed of a neural network.

12. An information processing device according to claim 11, wherein the detection unit detects the presence or absence of damage to the inspection object using the trained model to which learning data consisting of images relating to damage to the inspection object is input as parameters.

13. An information processing device according to claim 12, wherein the detection unit detects the presence or absence of damage to the inspection object based on the entire image.

14. An information processing method in which a computer system executes the following steps: generating an overall image based on event data acquired from an EVS (Event-based Vision Sensor); detecting an inspection object based on the overall image; and calculating position information of the inspection object based on the detection result by the detection unit.

15. An inspection system comprising: an EVS (Event-based Vision Sensor); an image generation unit that generates an overall image based on event data acquired from the EVS; a detection unit that detects an inspection object based on the overall image; and an information processing device having a calculation unit that calculates position information of the inspection object based on the detection result by the detection unit.

16. An inspection system as described in claim 15, wherein the information processing device has a control unit that controls a partial camera capable of capturing an image of an area corresponding to the position information of the inspection object based on the position information, and the partial camera is multiple.

Images