Pedestrian overpass diagnosis system, pedestrian overpass diagnosis method, and recording medium

Abstract

This pedestrian overpass diagnosis system comprises an acquisition unit, a detection unit, an estimation unit, a determination unit, and an output unit. The acquisition unit acquires an image which is of a pedestrian overpass that is placed over a road and which is captured using a device that is installed in a mobile body. The detection unit detects a portion of the pedestrian overpass in the acquired image. The estimation unit estimates the soundness of the portion of the pedestrian overpass on the basis of the image of the detected portion of the pedestrian overpass. The determination unit determines a response plan in repairs to the pedestrian overpass on the basis of the estimated soundness of the portion of the pedestrian overpass. The output unit outputs the determined response plan.

Description

Footbridge diagnostic system, footbridge diagnostic method, and recording medium

[0001] The present disclosure relates to a pedestrian bridge diagnostic system and the like.

[0002] For example, a road administrator checks the condition of a footbridge over a road by patrolling the road, and determines whether or not the footbridge needs repair based on the condition of the footbridge.

[0003] The structure inspection support system of Patent Document 1 acquires photographed images of a structure as monitoring data, and creates an inspection and repair plan for the structure based on the results of analyzing the structure using the monitoring data.

[0004] Japanese Patent Application Laid-Open No. 2019-57192

[0005] With the technology described in Patent Document 1, it may be difficult to accurately determine a response policy for a pedestrian bridge.

[0006] In order to solve the above-mentioned problems, the present disclosure aims to provide a pedestrian bridge diagnostic system etc. that can improve the accuracy of determining response policies for pedestrian bridge repairs.

[0007] In order to solve the above problems, the pedestrian bridge diagnosis system disclosed herein comprises an acquisition means for acquiring images of a pedestrian bridge installed on a road, taken using a device mounted on a mobile body; a detection means for detecting the parts of the pedestrian bridge that appear in the acquired images; an estimation means for estimating the soundness of the parts of the pedestrian bridge based on the images of each detected part of the pedestrian bridge; a decision means for determining a response policy for repairing the pedestrian bridge based on the estimated soundness of the parts of the pedestrian bridge; and an output means for outputting the determined response policy.

[0008] The pedestrian bridge diagnosis method disclosed herein acquires an image of a pedestrian bridge installed on a road taken using a device mounted on a mobile body, detects the parts of the pedestrian bridge that appear in the acquired image, estimates the health of each part of the pedestrian bridge based on the images of the detected pedestrian bridge parts, determines a response policy for repairing the pedestrian bridge based on the estimated health of the pedestrian bridge parts, and outputs the determined response policy.

[0009] The recording medium of the present disclosure non-temporarily records a pedestrian bridge diagnosis program that causes a computer to execute the following processes: acquiring images of a pedestrian bridge installed on a road photographed using a device mounted on a mobile body; detecting the parts of the pedestrian bridge that appear in the acquired images; estimating the soundness of the parts of the pedestrian bridge based on the images of each of the detected parts of the pedestrian bridge; determining a response policy for repairing the pedestrian bridge based on the estimated soundness of the parts of the pedestrian bridge; and outputting the determined response policy.

[0010] According to the present disclosure, it is possible to improve the accuracy of determining a response policy for repairing a pedestrian bridge.

[0011] FIG. 1 is a diagram illustrating an example of the configuration of a road management system according to an embodiment of the present disclosure. FIG. 2 is a diagram illustrating an example of the configuration of a pedestrian bridge diagnostic system according to an embodiment of the present disclosure. FIG. 3 is a diagram illustrating an example of a health level determination category according to an embodiment of the present disclosure. FIG. 4 is a diagram illustrating an example of a display screen according to an embodiment of the present disclosure. FIG. 5 is a diagram illustrating an example of a display screen according to an embodiment of the present disclosure. FIG. 6 is a diagram illustrating an example of a display screen according to an embodiment of the present disclosure. FIG. 7 is a diagram illustrating an example of an operation flow of a pedestrian bridge diagnostic system according to an embodiment of the present disclosure. FIG. 8 is a diagram illustrating an example of a hardware configuration according to an embodiment of the present disclosure.

[0012] An embodiment of the present disclosure will be described in detail with reference to the drawings. FIG. 1 is a diagram illustrating an example of the configuration of a road diagnosis system. The road diagnosis system includes a pedestrian bridge diagnosis system 10, an on-board device 20, and a terminal device 30. The pedestrian bridge diagnosis system 10 is connected to the on-board device 20 via a network. The pedestrian bridge diagnosis system 10 is also connected to the terminal device 30 via the network. Data input / output between the pedestrian bridge diagnosis system 10 and the on-board device 20 may be performed via a storage device. For example, data input / output between the pedestrian bridge diagnosis system 10 and the on-board device 20 may be performed via a non-volatile semiconductor storage device. There may be a plurality of on-board devices 20 and a plurality of terminal devices 30. The number of on-board devices 20 and the number of terminal devices 30 are set as appropriate.

[0013] The road diagnosis system is, for example, a system that determines a response policy for repairing a pedestrian bridge. The response policy for repairing a pedestrian bridge is, for example, whether or not the pedestrian bridge needs repair. That is, the road diagnosis system is, for example, a system that determines whether or not the pedestrian bridge needs repair. The response policy may be the timing of repair, the priority of repair, the urgency of repair, or the content of repair. The response policy may also be a combination of one or more items from among the necessity of repair, the timing of repair, the priority of repair, the urgency of repair, and the content of repair. The response policy is not limited to the above. The repair may also include the removal of the pedestrian bridge. The pedestrian bridge is, for example, a facility installed to allow pedestrians to cross a road. The pedestrian bridge may include an elevated pedestrian road. An elevated pedestrian road is, for example, a pedestrian deck. The pedestrian bridge is not limited to the above.

[0014] The road diagnosis system determines a course of action for repairing a pedestrian bridge based on, for example, images of a road taken using a camera mounted on a mobile object. The images may include moving images. The mobile object may be, for example, a vehicle. The mobile object may also be an unmanned aerial vehicle. However, the mobile object is not limited to the above.

[0015] When the moving body is a vehicle, the in-vehicle device 20 is, for example, a drive recorder. The in-vehicle device 20 is not limited to a drive recorder. Furthermore, the vehicle may be, for example, a vehicle used for road monitoring by a road administrator who manages a pedestrian bridge. The vehicle may also be a vehicle owned by a party other than the road administrator. For example, the vehicle may be a route bus, a tourist bus, a taxi, a freight vehicle, or a public vehicle. The vehicle may also include a motorcycle. For example, the road administrator uses a vehicle equipped with the in-vehicle device 20 to travel on a road on which a pedestrian bridge under management is installed. The road administrator then determines whether or not the pedestrian bridge needs to be repaired, for example, by referring to the response policy decision result output by the pedestrian bridge diagnosis system 10. Furthermore, when the moving body is not a vehicle, the in-vehicle device 20 may include a device equipped with a photographing function that is mounted on the moving body.

[0016] Here, an example of the configuration of the pedestrian bridge diagnostic system 10 will be described. Fig. 2 is a diagram showing an example of the configuration of the pedestrian bridge diagnostic system 10. The pedestrian bridge diagnostic system 10 basically comprises an acquisition unit 11, a detection unit 12, an estimation unit 13, a determination unit 14, and an output unit 16. The pedestrian bridge diagnostic system 10 may further comprise, for example, a prediction unit 15 and a storage unit 17.

[0017] The acquisition unit 11 acquires images of a pedestrian bridge installed on a road, captured using a device mounted on a mobile object. As images showing a pedestrian bridge installed on a road, for example, the acquisition unit 11 acquires moving images captured using a camera mounted as an in-vehicle device 20 on a vehicle traveling on the road. The moving images may include frames in which the pedestrian bridge is not captured. For example, the moving images captured using a camera mounted on a vehicle are moving images captured by a drive recorder.

[0018] The acquisition unit 11 acquires, for example, an image captured in front of a vehicle traveling on a road. The front of the vehicle is the direction in which the vehicle travels during normal driving. For example, the front of the vehicle is the direction that faces an occupant when the occupant sits in the driver's seat. The acquisition unit 11 may acquire an image captured in the rear of the vehicle. The acquisition unit 11 may also acquire an image captured in the left or right direction relative to the front of the vehicle. The left and right directions may include directions oblique to the front. The acquisition unit 11 may also acquire an image captured in an upward direction relative to the road surface. The upward direction relative to the road surface refers to a direction above the direction parallel to the road surface. The acquisition unit 11 may also acquire images captured in multiple directions. For example, the acquisition unit 11 may acquire images from multiple imaging devices mounted on the vehicle for each imaging direction.

[0019] The acquisition unit 11 may acquire location information of the moving object at the time of capturing the image together with the image showing the pedestrian bridge. The location information of the moving object is identified by the in-vehicle device 20 using the Global Navigation Satellite System (GNSS), for example. The acquisition unit 11 may also acquire an image to which the location information at the time of capturing the image has been added.

[0020] When the part that is the target of determining a response policy is selectable, the acquisition unit 11 may acquire a selection result of the part of the pedestrian bridge that is the target of determining a response policy. The part of the pedestrian bridge that is the target of determining a response policy is input, for example, by the road administrator to the terminal device 30. The acquisition unit 11 acquires, for example, from the terminal device 30, the part of the pedestrian bridge that is the target of determining a response policy that is input by the road administrator as the selection result.

[0021] The detection unit 12 detects parts of the pedestrian bridge that appear in the acquired image. Detecting parts of the pedestrian bridge means, for example, detecting each part of the pedestrian bridge from the image that shows the pedestrian bridge. For example, the detection unit 12 detects areas in which each part of the pedestrian bridge appears in the acquired image. The detection unit 12 may detect parts from the image that shows the pedestrian bridge that are targets for determining a repair policy for each part. For example, if there is only one part that is targets for determining a repair policy, the detection unit 12 may detect only that one part that is targets for determining a repair policy. The detection unit 12 detects, for example, a superstructure, a substructure, and a bridge girder as parts of the pedestrian bridge. The superstructure is, for example, a part above the pedestrian bridge where pedestrians cross the road. In other words, the superstructure is, for example, the bridge girder portion of the pedestrian bridge. The substructure is, for example, a part that supports the superstructure. In other words, the subwork is, for example, a support portion that holds the superstructure. The stairs are, for example, a portion for pedestrians to walk on between the walking portion of the superstructure and the sidewalk. The detection unit 12 may further detect accessories attached to the pedestrian bridge. Examples of accessories include lighting equipment and road signs. However, the accessories are not limited to the above.

[0022] The detection unit 12 detects the parts of the pedestrian bridge using, for example, a detection model. The detection model is, for example, a machine learning model that detects the parts of the pedestrian bridge from an image that shows the pedestrian bridge. The detection model may also detect areas in which each part of the pedestrian bridge is shown from an image. The detection unit 12 may detect the parts of the pedestrian bridge using two types of detection models: a detection model that detects the pedestrian bridge from an image, and a detection model that detects each part of the pedestrian bridge. The detection model is generated, for example, by learning the relationship between the image that shows each part of the pedestrian bridge and the name of the part. The detection model may also be generated by learning the relationship between the image that shows each part of the pedestrian bridge and the name of the part. The detection model is generated, for example, by deep learning using a neural network. The detection model is also generated, for example, in a system external to the pedestrian bridge diagnosis system 10.

[0023] The detection unit 12 may detect more detailed parts for each part. A detailed part refers to a smaller part that makes up a part. For example, in a superstructure, the detection unit 12 may detect the main girders and handrails that make up the superstructure. The unit by which the detection unit 12 detects parts may be set as appropriate. Furthermore, the parts that the detection unit 12 detects are not limited to those described above.

[0024] The detection unit 12 may detect an area in which a part is shown in each of images taken of the same pedestrian bridge from at least two directions. For example, the detection unit 12 detects the part of the pedestrian bridge from each of images taken of the same pedestrian bridge during multiple runs. For example, if a pedestrian bridge is installed between point A and point B, the detection unit 12 detects the part of the pedestrian bridge from an image taken when driving from point A to point B and an image taken when driving from point B to point A.

[0025] Furthermore, the detection unit 12 may detect each portion of the pedestrian bridge in images taken from each of the roads that intersect at the location where the pedestrian bridge is installed. For example, the detection unit 12 acquires images taken from each of the roads that extend perpendicular to the longitudinal direction of the superstructure of the pedestrian bridge and the roads that extend parallel to the longitudinal direction of the superstructure. For example, suppose there is an intersection where a road connecting points A and B and a road connecting points C and D intersect, and a pedestrian bridge is installed at the intersection so that the road connecting points A and B can be crossed. The detection unit 12 may detect the portion of the pedestrian bridge from, for example, an image taken when traveling from point A to point B and an image taken when traveling from point C to point D.

[0026] The detection unit 12 may also detect the portion of a pedestrian bridge from an image captured when turning right or left at an intersection. For example, suppose there is an intersection where a road connecting points A and B intersects with a road connecting points C and D, and a pedestrian bridge is installed on the point A side of the intersection so that the road connecting points A and B can be crossed. In this case, the detection unit 12 detects the portion of the pedestrian bridge from an image captured when turning left or right toward point A at the intersection while traveling from point C to point D. In this case, the detection unit 12 may also detect the portion of the pedestrian bridge from an image captured while traveling straight on the road connecting points A and B.

[0027] When detecting a pedestrian bridge portion from a video, the detection unit 12 may identify frames in which a pedestrian bridge may appear, and perform processing to detect the pedestrian bridge portion from the identified frames. For example, the detection unit 12 refers to location information at the time of shooting, and performs processing to detect the pedestrian bridge portion from frames in which a pedestrian bridge may appear. In this case, the detection unit 12 refers to, for example, a table that associates the location of a pedestrian bridge with identification information of the pedestrian bridge, and identifies images of frames in which a pedestrian bridge may appear.

[0028] The detection unit 12, for example, performs a process of associating a pedestrian bridge shown in an image with identification information of the pedestrian bridge. For example, when multiple pedestrian bridges are shown in a video, the detection unit 12 performs a process of associating identification information for each pedestrian bridge shown in the image. The identification information may include the location where the pedestrian bridge is installed. The detection unit 12 identifies the location shown in the image, for example, based on location information at the time the image was captured. The detection unit 12 then identifies the pedestrian bridge by referring to a table that associates the location where the pedestrian bridge is installed with the identification information of the pedestrian bridge. For example, the detection unit 12 associates the identified identification information of the pedestrian bridge with an image in which a pedestrian bridge portion is detected, and stores the image. The identification information of the pedestrian bridge is, for example, a number assigned to each pedestrian bridge. The identification information of the pedestrian bridge is not limited to the above.

[0029] The detection unit 12 may identify a pedestrian bridge based on characters written on the pedestrian bridge. The detection unit 12, for example, detects a place name written on a superstructure. For example, the detection unit 12 identifies a pedestrian bridge from the detected place name using a data table that associates place names with pedestrian bridge identification numbers. The detection unit 12 may also identify a pedestrian bridge based on a notice attached to the pedestrian bridge. The detection unit 12 may also identify a pedestrian bridge based on a notice that appears in the same image as the pedestrian bridge. For example, the detection unit 12 may identify a pedestrian bridge based on a notice attached to a traffic light at an intersection where the pedestrian bridge is installed. The detection unit 12, for example, detects a place name or intersection name written on the notice. Then, the detection unit 12 identifies a pedestrian bridge from the detected place name or intersection name using a data table that associates place names or intersection names with pedestrian bridge identification numbers. The detection unit 12 identifies the characters written on the pedestrian bridge using, for example, an identification model that identifies characters written on pedestrian bridges from an image. The detection unit 12 then refers to a table that associates characters written on the pedestrian bridge with identification information of the pedestrian bridge, and identifies the pedestrian bridge shown in the image based on the identified characters. Character identification may also be performed using a detection model that detects parts of the pedestrian bridge.

[0030] The estimation unit 13 estimates the soundness of each part of the pedestrian bridge based on the image in which the parts of the pedestrian bridge are detected. The soundness is, for example, an index indicating the degree of soundness of the pedestrian bridge and each part of the pedestrian bridge. For example, high soundness means, for example, that the parts of the pedestrian bridge are not deteriorating and the pedestrian bridge is in good condition. In other words, high soundness means that there is little need for repairs. Examples of deterioration of parts of the pedestrian bridge include cracks, discoloration, and corrosion. Deterioration of parts of the pedestrian bridge is not limited to the above. For example, the value indicating the soundness increases as the soundness increases. In other words, the value indicating the soundness decreases as the pedestrian bridge deteriorates. For example, the soundness may be an index whose value decreases as the soundness increases.

[0031] The estimation unit 13 may estimate the soundness of each part of the pedestrian bridge for each deterioration type. The deterioration type is a classification indicating the type of deterioration of the part of the pedestrian bridge. The deterioration type is, for example, cracking, discoloration, or corrosion. The deterioration type is not limited to the above. For example, the estimation unit 13 may estimate the soundness of cracking, discoloration, and corrosion. The estimation unit 13 may also estimate the soundness of the part of the pedestrian bridge based on the soundness of each deterioration type. For example, the estimation unit 13 converts the soundness of each of cracking, discoloration, and corrosion into a score. Then, the estimation unit 13 estimates a score indicating the soundness of the part of the pedestrian bridge based on the score indicating the soundness of each of cracking, discoloration, and corrosion. The estimation unit 13 may also weight the score for each deterioration type to estimate the score indicating the soundness of the part of the pedestrian bridge.

[0032] Furthermore, when more detailed parts are detected for each part, the estimation unit 13 may estimate the soundness of each of the detailed parts. For example, when the main girder and handrail of the superstructure are detected, the estimation unit 13 may estimate the soundness of each of the main girder and handrail of the superstructure. Furthermore, the estimation unit 13 may estimate the soundness of a part composed of the detailed parts based on the soundness of the detailed parts.

[0033] The estimation unit 13 estimates the health of each part shown in the image acquired by the acquisition unit 11 using, for example, an estimation model. The estimation model is, for example, a machine learning model that estimates the health of each part of a pedestrian bridge from an image showing the part. The estimation unit 13 estimates the health of each part, for example, by using, as input to the estimation model, an image of a region showing each part of the pedestrian bridge among the images showing the pedestrian bridge. The estimation unit 13 may estimate the health of each part by using, as input to the estimation model, an image of the pedestrian bridge with information indicating the region showing each part added thereto. Alternatively, the estimation unit 13 may estimate the health of each part by using, as input to the estimation model, an image of the pedestrian bridge in which a region not showing the part whose health is to be estimated among the images showing the pedestrian bridge is masked. The estimation model is generated, for example, by learning the relationship between the image showing each part and the health. The estimation model is generated, for example, by deep learning using a neural network. The estimation model is generated, for example, in a system external to the pedestrian bridge diagnostic system 10.

[0034] The estimation unit 13 may estimate the soundness of parts of the pedestrian bridge using estimation models corresponding to parts of the pedestrian bridge. For example, the estimation unit 13 estimates the soundness of each part using an estimation model for the superstructure, an estimation model for the substructure, and an estimation model for the stairs. For example, the estimation unit 13 uses an image of the superstructure as input to the estimation model for the superstructure to estimate the soundness of the superstructure. Furthermore, the estimation unit 13 uses an image of the substructure as input to the estimation model for the substructure to estimate the soundness of the substructure. Furthermore, the estimation unit 13 uses an image of the stairs as input to the estimation model for the stairs to estimate the soundness of the stairs. Furthermore, the classification of the estimation models is not limited to the above.

[0035] The estimation unit 13 may estimate the soundness of each part of a pedestrian bridge based on images taken from at least two directions. For example, when a pedestrian bridge is installed between point A and point B, the estimation unit 13 estimates the soundness of the superstructure by using, for example, an image of the superstructure taken when traveling from point A to point B and an image of the superstructure taken when traveling from point A to point B as inputs to an estimation model. For example, the estimation unit 13 estimates, for each part of the pedestrian bridge, the result of the soundness in the direction with the lowest soundness as the soundness of each part. The estimation unit 13 may also estimate, for each part of the pedestrian bridge, the result of the soundness in the direction with the highest soundness as the soundness of each part. Furthermore, the estimation unit 13 may estimate, for each part of the pedestrian bridge, the average value of the soundness in each direction as the soundness of each part.

[0036] The estimation unit 13 may estimate the health of the parts shown in the image based on changes over time in the parts shown in the image. For example, the estimation unit 13 estimates the health of the parts of the pedestrian bridge based on a reference image and the image acquired by the acquisition unit 11. The reference image is an image taken when the pedestrian bridge is repaired or installed. The reference image may be an image taken at a time other than when the pedestrian bridge is repaired or installed. The reference image may also be multiple images. When there are multiple reference images, the images may be taken at different times. The estimation unit 13 estimates the health of each part, for example, using the reference image and the image whose health is to be estimated as inputs to an estimation model. In this case, the estimation unit 13 estimates the health of the parts of the pedestrian bridge using, for example, an estimation model that estimates the health from the difference between the reference image and the image whose health is to be estimated. An estimation model that estimates the health level from the difference between two images is generated, for example, by learning the relationship between the reference image, the image whose health level is to be estimated, and the health level.

[0037] The estimation unit 13 may estimate the healthiness of each part of the pedestrian bridge based on images showing each part of the pedestrian bridge and the environment in which the pedestrian bridge is installed. In this case, the estimation unit 13, for example, inputs the images showing each part of the pedestrian bridge and the environment in which the pedestrian bridge is installed, and estimates the healthiness of each part using an estimation model that estimates the healthiness of each part. The environment in which the pedestrian bridge is installed is, for example, an environment that may affect the progression of deterioration of the pedestrian bridge. The environment in which the pedestrian bridge is installed is, for example, one or more of the amount of precipitation, amount of snowfall, amount of road traffic, atmospheric composition, hours of sunshine, direction of sunlight on the pedestrian bridge, and distance from the sea. The environment in which the pedestrian bridge is installed is not limited to the above.

[0038] For example, even if the appearances are similar, the soundness of each part may differ depending on the surrounding conditions. For example, even if the appearances are similar, if deterioration is progressing due to chemical substances or salt in the atmosphere, the estimation unit 13 estimates the soundness of the part to be estimated so that the soundness is lower than that of a pedestrian bridge installed in a location where there are no factors that cause deterioration to progress.

[0039] The estimation unit 13 may estimate the soundness of each part further based on the shape of the part of the footbridge. For example, the estimation unit 13 estimates the soundness of each part using an estimation model that estimates the soundness of each part, using an image showing each part of the footbridge and the shape of each part of the footbridge as input to the estimation model.

[0040] The estimation unit 13 may also estimate the soundness of the pedestrian bridge based further on the shape of the pedestrian bridge's exterior. In this case, the estimation unit 13, for example, inputs an image showing each part of the pedestrian bridge and the shape of the pedestrian bridge's exterior, and estimates the soundness of each part using an estimation model that estimates the soundness of each part. The shape of the pedestrian bridge's exterior is, for example, the shape of the superstructure. For example, the shape of the superstructure is the shape of the pedestrian bridge's appearance when viewed from above. The shape of the pedestrian bridge when viewed from above may be, for example, a circular shape or a shape consisting of a combination of straight lines. The shape of the pedestrian bridge's exterior may also be the shape of stairs. For example, the shape of the stairs may be straight or have a turn in the middle. The shape of the pedestrian bridge's exterior is not limited to the above. For example, a pedestrian bridge with a straight superstructure and a pedestrian bridge with a circular superstructure may have different soundness levels even if their appearances are similar. Therefore, further using the shape of the pedestrian bridge's exterior may improve the accuracy of the soundness estimation.

[0041] The estimation unit 13 may estimate the soundness of a portion of a pedestrian bridge not visible in an image based on the soundness of multiple portions. For example, the estimation unit 13 estimates the soundness of the deck slab of the superstructure based on the soundness of the stairs and the soundness of the superstructure. For example, in an image of a pedestrian bridge, the main girder of the superstructure and the walkway of the stairs may be visible in the image. On the other hand, the walkway on the top surface of the superstructure cannot be photographed from below the pedestrian bridge. Therefore, the estimation unit 13 estimates the soundness of the walkway of the superstructure using the soundness of the main girder portion of the superstructure and the soundness of the walkway of the stairs, which is also a walkway. The relationship between the soundness of multiple portions and the soundness of the portion to be estimated is set, for example, as data in a table. Furthermore, the estimation unit 13 may estimate the soundness of the portion not visible in an image using a learning model that estimates the soundness of the portion not visible in an image from the soundness of multiple portions.

[0042] The estimation unit 13 may estimate the healthiness of the entire pedestrian bridge based on the healthiness of each part of the pedestrian bridge. For example, the estimation unit 13 estimates the value obtained by multiplying the healthiness of each part of the pedestrian bridge by a weight and adding the results as the healthiness of the entire pedestrian bridge. The weight of each part is set, for example, so that the more important the part, the larger the weight value. The weight of each part is set, for example, by the road administrator. The estimation unit 13 may estimate the healthiness of the entire pedestrian bridge based on a predetermined statistical value of the healthiness of each part of the pedestrian bridge. The predetermined statistical value is, for example, an average value, a median value, a maximum value, or a minimum value. The predetermined statistical value is not limited to the above.

[0043] The determination unit 14 determines a response policy for repairing the pedestrian bridge based on the estimated soundness of the parts of the pedestrian bridge. The determination unit 14 determines a response policy for repairing the pedestrian bridge shown in the image, for example, from among the categories of response policies that have been set. The response policy for repairing the pedestrian bridge is, for example, a response policy for each part or a response policy for the entire pedestrian bridge.

[0044] The determination unit 14 determines whether or not each part of the pedestrian bridge needs to be repaired, for example, based on the soundness of each part of the pedestrian bridge, as a response policy for repairing the pedestrian bridge. For example, the determination unit 14 determines that parts whose soundness is below a standard are parts that need repair. The standard for soundness is set, for example, so that if the standard is met, the pedestrian bridge can continue to be operated without repair. The standard for soundness is set, for example, by the road administrator.

[0045] The determination unit 14 may determine a response policy for repairing the pedestrian bridge based on multiple levels of criteria. For example, the determination unit determines the urgency of repair based on the level to which the soundness level corresponds among the multiple levels. The urgency of repair is, for example, an indicator that indicates the need for early correction in order to safely operate the pedestrian bridge. For example, the urgency of repair is high when there is a possibility of an accident occurring if repair is not performed immediately. On the other hand, the urgency of repair is low when deterioration is progressing slowly, such as discoloration of the paint. The determination unit 14 determines a response policy for each part based on the urgency of each part, for example. Furthermore, the determination unit 14 may determine a response policy for the entire pedestrian bridge based on the part with the highest level of urgency.

[0046] The determination unit 14 may, for example, determine a category indicating a response policy as a response policy for repairing a pedestrian bridge. The category indicating a response policy is, for example, a category based on the urgency of repair. The determination unit 14 may, for example, determine the category of the response policy for each part of the pedestrian bridge based on the soundness of each part of the pedestrian bridge. Furthermore, the determination unit 14 may determine the category of the response policy for the pedestrian bridge as a whole based on the soundness of each part of the pedestrian bridge or the soundness of the pedestrian bridge as a whole.

[0047] FIG. 3 shows an example of classification of response policies for repairing a pedestrian bridge. In the example of classification in FIG. 3, a "decision classification," which is an identifier of the response policy classification, is associated with a "response content," which indicates the content of the response policy. In the example of classification in FIG. 3, for example, decision classification A is a classification in which the integrity is high and no response is necessary at the time of determination. Also, in the example of classification in FIG. 3, for example, decision classification C1 is a classification in which there are no safety issues at the time of determination, but repair is necessary for preventive maintenance. In the example of classification in FIG. 3, the determination result of decision classification C1 is determined, for example, when the paint has discolored and the protective effect of the paint has decreased, and it is desirable to paint. Also, in the example of classification in FIG. 3, for example, decision classification E1 is a classification in which, at the time of determination, safety issues may arise if repairs are not performed. In the example of classification in FIG. 3, the determination result of decision classification E1 is determined, for example, when damage has occurred to the stairs and immediate repairs are necessary.

[0048] The determination unit 14 may also determine the priority of repair for the pedestrian bridge. For example, the determination unit 14 determines that the more portions requiring repair, the higher the repair priority of the pedestrian bridge. The determination unit 14 may also determine whether or not repair is necessary based on a score indicating the soundness of the pedestrian bridge. For example, the determination unit 14 calculates a score related to the soundness of the entire pedestrian bridge based on the soundness of each portion. Then, the determination unit 14 determines whether or not repair is necessary for the pedestrian bridge based on, for example, the calculated score related to the soundness of the entire pedestrian bridge. When determining whether or not repair is necessary based on the score indicating the soundness of the pedestrian bridge, the determination unit 14 may calculate the score related to the soundness of the pedestrian bridge by weighting the score indicating the soundness of each portion. The weight of the soundness of each portion is set, for example, by the road administrator.

[0049] The determination unit 14 may determine a response policy for repairing the pedestrian bridge based on the health of the parts of the pedestrian bridge and the environment in which the pedestrian bridge is installed. For example, the determination unit 14 determines, as a response policy for repairing the pedestrian bridge, whether or not repair is necessary for the parts of the pedestrian bridge based on the health of each part and a score indicating the environment in which the pedestrian bridge is installed. The determination unit 14 determines whether or not repair is necessary for each part based on, for example, a value obtained by multiplying the reciprocal of the health of each part by the score indicating the environment in which the pedestrian bridge is installed. In this case, the score indicating the environment in which the pedestrian bridge is installed is set to a higher score as the influence on deterioration of the parts of the pedestrian bridge becomes stronger. The environment in which the pedestrian bridge is installed is, for example, an environment that can affect the progression of deterioration of the pedestrian bridge. The environment in which the pedestrian bridge is installed is, for example, one or more of the amount of precipitation, the amount of snowfall, the amount of traffic on the road, the atmospheric composition, the hours of sunshine, the direction of sunlight shining on the pedestrian bridge, and the distance from the sea. The environment in which the pedestrian bridge is installed is not limited to the above.

[0050] As a response policy for repairing the pedestrian bridge, the determination unit 14 may determine the priority of repairing the pedestrian bridge based on the soundness of each part and a score indicating the environment in which the pedestrian bridge is installed. For example, the determination unit 14 determines the priority of repairing the pedestrian bridge based on a value obtained by multiplying the reciprocal of the soundness of each part by the score indicating the environment in which the pedestrian bridge is installed. In this case, the score indicating the environment in which the pedestrian bridge is installed is set to a higher score as the impact on deterioration of the pedestrian bridge becomes stronger.

[0051] The determination unit 14 may determine a response policy for repairing a pedestrian bridge further based on the importance of the pedestrian bridge. The importance of a pedestrian bridge is, for example, an index that indicates the social importance of the pedestrian bridge. For example, the importance of a pedestrian bridge increases as the volume of traffic on the road increases. Furthermore, the importance of a pedestrian bridge increases as the number of people crossing the road increases. For example, the importance of a pedestrian bridge increases when a school, a children's facility, a hospital, a sports facility, an event facility, a commercial facility, a station, and a public facility are present. The determination unit 14 determines whether or not repairing a pedestrian bridge is necessary based on, for example, a value obtained by multiplying the reciprocal of the soundness of each part by the importance of the pedestrian bridge. The determination unit 14 may determine the priority for repairing a pedestrian bridge based on, for example, a value obtained by multiplying the reciprocal of the soundness of each part by the importance of the pedestrian bridge.

[0052] The prediction unit 15 predicts, for example, when repairs will be required for a pedestrian bridge. The prediction unit 15 also predicts, for example, when repairs will be required for a pedestrian bridge. The times when repairs will be required may include the time when removal will be required. The prediction unit 15 predicts, for example, when repairs will be required for each part based on time-series data on the health of each part. For example, the prediction unit 15 predicts when the health will fall below a standard. Then, the prediction unit 15 predicts when the health of each part will fall below the standard as the time when repairs will be required for each part. The time to remove a pedestrian bridge may also be predicted based on when the health of multiple parts will fall below the standard.

[0053] The prediction unit 15 may predict when repairs will be required for the pedestrian bridge using a prediction model that predicts when repairs will be required from time-series data indicating the health of the pedestrian bridge parts. The prediction model is generated, for example, by learning the relationship between the time-series data indicating the health of the pedestrian bridge parts and the time it takes for repairs to be required. The prediction model may also predict when repairs will be required from the health of each pedestrian bridge part and the environment in which the pedestrian bridge is installed. In this case, the prediction model is generated, for example, by learning the relationship between the health of each pedestrian bridge part, the environment in which the pedestrian bridge is installed, and the time until repairs will be required.

[0054] The prediction unit 15 may predict the details of necessary repairs based on the soundness level for each deterioration type. For example, if the soundness levels for discoloration and rust are low, the prediction unit 15 predicts that painting is necessary as the necessary repair. In this case, the prediction unit 15 predicts the details of necessary repairs using a prediction model that predicts the details of necessary repairs from the soundness level for each deterioration type. The prediction model is generated by learning the relationship between the soundness level for each deterioration type and the details of necessary repairs. The prediction model is generated, for example, in a system external to the pedestrian bridge diagnosis system 10.

[0055] The prediction unit 15 may predict a location where it is desirable to install a pedestrian bridge. The prediction unit 15 predicts a location where it is desirable to install a pedestrian bridge, for example, based on the predicted time when an installed pedestrian bridge will need to be removed and the importance of the pedestrian bridge at each location. The prediction unit 15 may also estimate a location where it is desirable to install a new pedestrian bridge while leaving an existing pedestrian bridge in place. The importance of the pedestrian bridge is, for example, an index based on the number of people predicted to cross a road and the traffic volume on the road. For example, when it is predicted that a pedestrian bridge will be removed, the prediction unit 15 predicts a location where it is desirable to install a pedestrian bridge based on the number of people predicted to cross the road at each location in the case where there is no pedestrian bridge to be removed. The number of people crossing the road is predicted based on, for example, the population of the area and surrounding facilities.

[0056] The output unit 16 outputs the determined response policy. The output unit 16 outputs, for example, the result of determining the response policy for each pedestrian bridge. The output unit 16 may output the result of determining the response policy for each pedestrian bridge using the classification of the response policy. The output unit 16 may output information about the portion determined to require repair and the pedestrian bridge to which the portion is attached. The output unit 16 may further output an image of the pedestrian bridge determined to require repair.

[0057] The output unit 16 may output the determination result of the response policy for each part of the pedestrian bridge. The output unit 16 may output the soundness of the part of the pedestrian bridge along with the determination result of the response policy. The output unit 16 may output at least one of the soundness and the response policy for each part of the pedestrian bridge. Furthermore, the output unit 16 may output information indicating pedestrian bridges that need repair in descending order of repair priority. Furthermore, when a prediction is made of the time when the pedestrian bridge will need repair, the output unit 16 further outputs, for example, a prediction result of the time when the pedestrian bridge will need repair.

[0058] The output unit 16 may output the response policy for repair of each pedestrian bridge by superimposing it on the map. For example, the output unit 16 outputs the response policy for each pedestrian bridge by superimposing it on the map using a display mode corresponding to the response policy. The display mode corresponding to the response policy is, for example, a symbol, character, or color set according to the response policy. The display mode corresponding to the response policy is not limited to the above. Furthermore, the output unit 16 may output the health level of each pedestrian bridge by superimposing it on the map using a display mode corresponding to the health level. The mode corresponding to the health level is a symbol, a numerical value, character, or color. The mode corresponding to the health level is not limited to the above. Furthermore, when a pedestrian bridge is selected on the map displayed on the screen, the output unit 16 may further output detailed information about the pedestrian bridge.

[0059] FIG. 4 is an example of a display screen showing the result of determining a response policy for repairing a footbridge. The output unit 16 outputs the result shown in the example of the display screen of FIG. 4 based on the result of determining the response policy. In the example of the display screen of FIG. 4, a "footbridge number," an "installation location," and a "determination category" are associated with each footbridge. In the example of the display screen of FIG. 4, the "footbridge number" is, for example, the identification number of the footbridge. In the example of the display screen of FIG. 4, the "installation location" is, for example, information indicating the location where the footbridge is installed. In the example of the display screen of FIG. 4, the "installation location" is, for example, indicated by the latitude and longitude and the name of the intersection where the footbridge is installed. In the example of the display screen of FIG. 4, if a footbridge is not installed at an intersection, the name of the intersection is displayed as a blank, for example. In the example of the display screen of FIG. 4, the "determination category" is, for example, a category of the response policy for repairing the sidewalk. A road administrator can determine whether or not repair is necessary for each footbridge by referring to the result of determining the response policy shown in the example of the display screen of FIG. 4.

[0060] FIG. 5 is an example of a display screen showing the results of the response policy determination for each part of a pedestrian bridge. The output unit 16 outputs the determination results shown in the example of the display screen of FIG. 5 based on, for example, the determination results for each part. In the example of the display screen of FIG. 5, a "pedestrian bridge number," an "installation location," and a "determination category" are associated with each pedestrian bridge. In the example of the display screen of FIG. 5, the "pedestrian bridge number" is, for example, the identification number of the pedestrian bridge. Also, in the example of the display screen of FIG. 5, the "installation location" is information indicating the location where the pedestrian bridge is installed in a format similar to the example of the display screen of FIG. 4. Also, in the example of the display screen of FIG. 5, the "superstructure," "substructure," and "stairs" indicate the results of the response policy determination for each part. For example, by referring to the result of the response policy determination shown in the example of the display screen of FIG. 5, a road administrator can determine whether or not repairs are necessary for each pedestrian bridge.

[0061] 6 is an example of a display screen that displays the result of the response policy determination for a pedestrian bridge superimposed on a map. The output unit 16 outputs the result of the response policy determination for a pedestrian bridge superimposed on a map, for example, as shown in the example of the display screen in FIG. 6. In the example of the display screen in FIG. 6, the response policies for "Pedestrian Bridge S," "Pedestrian Bridge T," and "Pedestrian Bridge U," respectively, are displayed as "A," "B," and "C1." For example, by referring to the result of the response policy determination shown in the example of the display screen in FIG. 6, the road administrator can easily identify pedestrian bridges that need repair.

[0062] FIG. 7 is an example of a display screen that displays detailed information about a pedestrian bridge when the pedestrian bridge is selected on a map. When, for example, "pedestrian bridge T" is selected on the display screen shown in the example of FIG. 6 , the output unit 16 outputs detailed information about the "pedestrian bridge T" by superimposing it on the map, as shown in the example of the display screen of FIG. 7 . In the example of the display screen of FIG. 7 , for example, detailed information about the "pedestrian bridge T" includes an image of the "pedestrian bridge T," its "latitude," its "longitude," its "type," its "determination category," its "travel date and time," its "status," and its "route name." In the example of the display screen of FIG. 7 , the "latitude" and "longitude" indicate, for example, the location of the pedestrian bridge by latitude and longitude. In the example of the display screen of FIG. 7 , the "type" indicates, for example, the shape of the pedestrian bridge. In the example of the display screen of FIG. 7 , the "determination category" indicates, for example, the classification of the response policy for the pedestrian bridge. In the example of the display screen of FIG. 7 , the "status" indicates, for example, the degree of deterioration of each part. The "status" in the example of the display screen of FIG. 7 indicates that there are moderate cracks and minor discoloration. In addition, in the example of the display screen in Fig. 7, "Route Name" is the name of the road on which the pedestrian bridge is installed. By referring to detailed information about the pedestrian bridge, such as that shown in the example of the display screen in Fig. 7, the road administrator can more accurately determine whether or not each pedestrian bridge needs repair.

[0063] The memory unit 17 stores data related to the process of determining a response policy for the pedestrian bridge. The memory unit 17 stores, for example, images acquired by the acquisition unit 11. The memory unit 17 stores, for example, detection results for parts of the pedestrian bridge. The memory unit 17 stores, for example, estimation results for the soundness of each part of the pedestrian bridge. The memory unit 17 stores, for example, determination results for a response policy for the pedestrian bridge. The memory unit 17 stores, for example, detection models. The memory unit 17 stores, for example, recognition models. The memory unit 17 stores, for example, prediction models. The detection models, recognition models, and prediction models may be stored in storage means other than the memory unit 17.

[0064] The in-vehicle device 20 is, for example, mounted on a moving body and includes a camera that captures an image in front of the moving body. The moving body is, for example, a vehicle. The camera of the in-vehicle device 20 captures an image of a pedestrian bridge installed on a road. The camera of the in-vehicle device 20 may also capture an image behind, to the left, or to the right of the vehicle. The in-vehicle device 20 may also include multiple camera devices.

[0065] For example, the in-vehicle device 20 adds information about the location where the video was captured to the captured video. For example, the in-vehicle device 20 uses GNSS to identify the location of the moving object at the time the video was captured. The in-vehicle device 20 may identify the location of the moving object based on a beacon containing location information. The in-vehicle device 20 may identify the capture location based on map information and the distance traveled from the location where the moving object's position was identified. Furthermore, the in-vehicle device 20 outputs the captured video and location information of the moving object at the time the video was captured to the acquisition unit 11 of the pedestrian bridge diagnosis system 10, for example. A drive recorder is used as the in-vehicle device 20. The in-vehicle device 20 is not limited to a drive recorder. For example, the in-vehicle device 20 may store the captured video and location information of the moving object at the time the video was captured in a removable storage medium.

[0066] The terminal device 30 is, for example, a terminal device used by a road administrator. The terminal device 30 outputs the result of the decision on a response policy for repairing a pedestrian bridge from, for example, the output unit 16 of the pedestrian bridge diagnosis system 10. The terminal device 30 outputs the acquired result of the decision on a response policy for repairing a pedestrian bridge to a display device (not shown).

[0067] When the part of the pedestrian bridge for which a response policy is to be determined can be selected, the terminal device 30 acquires the selection result of the part of the pedestrian bridge for which a response policy is to be determined, which is input by, for example, a road administrator. Then, the terminal device 30 outputs the selection result of the part of the pedestrian bridge for which a response policy is to be determined to, for example, the acquisition unit 11 of the pedestrian bridge diagnosis system 10.

[0068] Furthermore, when a pedestrian bridge for which detailed information is to be displayed can be selected on the map, the terminal device 30 acquires the selection result of the pedestrian bridge for which detailed information is to be displayed, which is input by, for example, a road administrator. The terminal device 30 then outputs the selection result of the pedestrian bridge for which detailed information is to be displayed to, for example, the acquisition unit 11 of the pedestrian bridge diagnosis system 10. Furthermore, the terminal device 30 can be, for example, a personal computer, a tablet computer, or a smartphone. The terminal device 30 is not limited to the above examples.

[0069] The following describes the operation of determining a course of action for repairing a pedestrian bridge in the pedestrian bridge diagnostic system 10. Fig. 8 shows an example of the flow of processing for determining a course of action for repairing a pedestrian bridge in the pedestrian bridge diagnostic system 10.

[0070] The acquisition unit 11 acquires an image of a pedestrian bridge installed on a road, which is captured using a device mounted on a moving object (step S11). The acquisition unit 11 acquires, for example, a video image captured while the vehicle is traveling on a road on which a pedestrian bridge is installed, from an in-vehicle device 20 mounted on the vehicle. The in-vehicle device 20 is, for example, a drive recorder mounted on the vehicle.

[0071] When an image showing a pedestrian bridge is acquired, the detection unit 12 detects the part of the pedestrian bridge that appears in the acquired image (step S12). The detection unit 12 detects the part of the pedestrian bridge that appears in the acquired image, for example, by using a detection model.

[0072] When the part of the footbridge is detected, the estimation unit 13 estimates the soundness of the part of the footbridge based on the image of the part of the footbridge detected by the detection unit 12 (step S13).

[0073] Once the soundness is estimated, the determination unit 14 determines a course of action for repairing the pedestrian bridge based on the estimated soundness of the part of the pedestrian bridge (step S14).

[0074] When the response policy has been determined for all pedestrian bridges shown in the acquired image (Yes in step S15), the output unit 16 outputs the response policy determined by the determination unit 14 (step S16).

[0075] In step S15, if there is a pedestrian bridge for which the response policy has not been determined (No in step S15), for example, the process returns to step S12, and the detection unit 12 detects parts of the pedestrian bridge from the image of the pedestrian bridge for which the response policy has not been determined.

[0076] The pedestrian bridge diagnostic system 10 acquires images of a pedestrian bridge installed on a road and detects each part of the pedestrian bridge that appears in the acquired images. The pedestrian bridge diagnostic system 10 estimates the soundness of the pedestrian bridge parts based on the detected images of the pedestrian bridge parts. The pedestrian bridge diagnostic system 10 then determines a response policy for repairing the pedestrian bridge based on the estimated soundness of the pedestrian bridge parts. In this way, by determining a response policy based on the estimated results of the soundness of the pedestrian bridge parts, it is possible to determine a response policy that is appropriate for the condition of the parts, and therefore the pedestrian bridge diagnostic system 10 can improve the accuracy of determining a response policy for the pedestrian bridge.

[0077] Furthermore, by determining a course of action for repairing a pedestrian bridge based on video images captured using the on-board device 20 mounted on a mobile body, the pedestrian bridge diagnostic system 10 can, for example, easily determine whether or not repairs to the pedestrian bridge are necessary. Furthermore, by determining a course of action for each part of the pedestrian bridge based on video images captured using the on-board device 20 mounted on a mobile body, the pedestrian bridge diagnostic system 10 can, for example, easily identify parts of the pedestrian bridge that require repairs.

[0078] Furthermore, by using the response policy classification to determine the response policy classification for a pedestrian bridge, for example, a road administrator can determine whether or not road repairs are necessary, taking into account the urgency of the repairs. Therefore, by using the pedestrian bridge diagnosis system 10, it is possible to more accurately determine whether or not road repairs are necessary.

[0079] Each process in the pedestrian bridge diagnostic system 10 may be distributed and executed among multiple information processing devices connected via a network. For example, the processes in the detection unit 12, estimation unit 13, and determination unit 14 and the process in the prediction unit 15 may be performed in different information processing devices. Also, for example, the processes in the detection unit 12 and estimation unit 13 and the processes in the determination unit 14 and prediction unit 15 may be performed in different information processing devices. It can be set as appropriate which of the multiple information processing devices performs each process in the pedestrian bridge diagnostic system 10.

[0080] Each process in the pedestrian bridge diagnostic system 10 can be realized by executing a computer program on a computer. Fig. 9 shows an example of the configuration of a computer 100 that executes a computer program that performs each process in the pedestrian bridge diagnostic system 10. The computer 100 includes a CPU (Central Processing Unit) 101, a memory 102, a storage device 103, an input / output I / F (Interface) 104, and a communication I / F 105.

[0081] The CPU 101 reads and executes computer programs for performing each process from the storage device 103. The CPU 101 may be configured with a combination of multiple CPUs. Furthermore, the CPU 101 may be configured with a combination of a CPU and another type of processor. For example, the CPU 101 may be configured with a combination of a CPU and a graphics processing unit (GPU). The memory 102 is configured with a dynamic random access memory (DRAM) or the like, and temporarily stores computer programs executed by the CPU 101 and data being processed. The storage device 103 stores computer programs executed by the CPU 101. The storage device 103 is configured with, for example, a non-volatile semiconductor storage device. Other storage devices such as a hard disk drive may also be used for the storage device 103. The input / output I / F 104 is an interface that receives input from an operator and outputs display data, etc. The communication I / F 105 is an interface that transmits and receives data between the in-vehicle device 20, the terminal device 30, and other information processing devices. Furthermore, the terminal device 30 may have a configuration similar to that of the computer 100.

[0082] The computer program used to execute each process can also be stored and distributed on a computer-readable recording medium that non-temporarily stores data. Examples of recording media that can be used include magnetic tapes for recording data and magnetic disks such as hard disks. Optical disks such as CD-ROMs (Compact Disc Read Only Memory) can also be used as recording media. Non-volatile semiconductor storage devices can also be used as recording media.

[0083] A part or all of the above-described embodiments can be described as, but not limited to, the following supplementary notes.

[0084] [Supplementary Note 1] A pedestrian bridge diagnostic system comprising: an acquisition means for acquiring images of a pedestrian bridge installed on a road, taken using a device mounted on a mobile body; a detection means for detecting parts of the pedestrian bridge that appear in the acquired images; an estimation means for estimating the soundness of the parts of the pedestrian bridge based on the detected images of the parts of the pedestrian bridge; a decision means for determining a course of action for repairing the pedestrian bridge based on the estimated soundness of the parts of the pedestrian bridge; and an output means for outputting the decided course of action.

[0085] [Supplementary Note 2] The pedestrian bridge diagnosis system according to Supplementary Note 1, wherein the estimation means determines a course of action for the pedestrian bridge based on the soundness of a plurality of parts of the pedestrian bridge.

[0086] [Appendix 3] A pedestrian bridge diagnostic system as described in Appendix 1 or 2, wherein the detection means detects parts of the pedestrian bridge in each of images taken of the same pedestrian bridge from at least two directions, and the estimation means estimates the soundness of each part of the pedestrian bridge based on the images of the parts of the pedestrian bridge taken from at least two directions.

[0087] [Supplementary Note 4] The pedestrian bridge diagnosis system according to Supplementary Note 3, wherein the detection means detects portions of the pedestrian bridge in images taken from each of the roads that intersect at the installation point of the pedestrian bridge.

[0088] [Supplementary Note 5] The pedestrian bridge diagnosis system according to any one of Supplementary Notes 1 to 4, wherein the estimation means estimates the soundness of the part of the pedestrian bridge based on a reference image and the image acquired by the acquisition means.

[0089] [Supplementary Note 6] The pedestrian bridge diagnostic system according to Supplementary Note 5, wherein the reference image is an image taken when the pedestrian bridge has been repaired or installed.

[0090] [Supplementary Note 7] The pedestrian bridge diagnostic system according to any one of Supplementary Notes 1 to 6, wherein the determination means determines a course of action for repairing the pedestrian bridge based on the soundness of parts of the pedestrian bridge and the environment in which the pedestrian bridge is installed.

[0091] [Supplementary Note 8] The pedestrian bridge diagnosis system according to any one of Supplementary Notes 1 to 7, wherein the estimation means estimates the soundness of the part of the pedestrian bridge based on an image of the part of the pedestrian bridge and the environment in which the pedestrian bridge is installed.

[0092] [Supplementary Note 9] The pedestrian bridge diagnostic system according to any one of Supplementary Notes 1 to 8, wherein the determining means determines a course of action for repairing the pedestrian bridge further based on the importance of the pedestrian bridge.

[0093] [Supplementary Note 10] A pedestrian bridge diagnostic system according to any one of Supplementary Notes 1 to 9, further comprising a prediction means for predicting when the pedestrian bridge will need to be repaired or removed, and the output means further outputs the prediction result of when the pedestrian bridge will need to be repaired or removed.

[0094] [Supplementary Note 11] The pedestrian bridge diagnosis system according to any one of Supplementary Notes 1 to 10, wherein the output means outputs the response policy for each of the pedestrian bridges by superimposing it on a map.

[0095] [Supplementary Note 12] The pedestrian bridge diagnostic system according to any one of Supplementary Notes 1 to 11, wherein the output means outputs the soundness of each of the pedestrian bridges by superimposing it on a map.

[0096] [Supplementary Note 13] The pedestrian bridge diagnosis system according to any one of Supplementary Notes 1 to 12, wherein the estimation means estimates the soundness further based on a shape of the pedestrian bridge.

[0097] [Supplementary Note 14] A pedestrian bridge diagnostic system according to any one of Supplementary Notes 1 to 13, wherein the estimation means estimates the soundness of each part of the pedestrian bridge shown in the image acquired by the acquisition means, using an estimation model that estimates the soundness of parts of the pedestrian bridge from the image showing the parts.

[0098] [Supplementary Note 15] The pedestrian bridge diagnosis system according to Supplementary Note 14, wherein the estimation means estimates the soundness of the part of the pedestrian bridge using an estimation model based on the part of the pedestrian bridge.

[0099] [Supplementary Note 16] The pedestrian bridge diagnosis system according to any one of Supplementary Notes 1 to 15, wherein the detection means identifies the pedestrian bridge based on characters written on the pedestrian bridge.

[0100] [Supplementary Note 17] The pedestrian bridge diagnostic system according to any one of Supplementary Notes 1 to 16, wherein the estimation means estimates the healthiness of a portion of the pedestrian bridge that is not shown in the image based on the healthiness of multiple portions of the pedestrian bridge.

[0101] [Supplementary Note 18] The pedestrian bridge diagnostic system according to any one of Supplementary Notes 1 to 17, wherein the images are time-series moving images captured using a device mounted on a vehicle.

[0102] [Supplementary Note 19] A pedestrian bridge diagnosis method comprising: acquiring an image of a pedestrian bridge installed on a road, photographed using a device mounted on a mobile body; detecting a portion of the pedestrian bridge that appears in the acquired image; estimating the soundness of the portion of the pedestrian bridge based on the detected image of the portion of the pedestrian bridge; determining a course of action for repairing the pedestrian bridge based on the estimated soundness of the portion of the pedestrian bridge; and outputting the determined course of action.

[0103] [Supplementary Note 20] A recording medium that non-temporarily records a pedestrian bridge diagnosis program that causes a computer to execute the following processes: acquiring an image of a pedestrian bridge installed on a road, taken using a device mounted on a moving object; detecting the parts of the pedestrian bridge that appear in the acquired image; estimating the soundness of the parts of the pedestrian bridge based on the detected image of the parts of the pedestrian bridge; deciding on a course of action for repairing the pedestrian bridge based on the estimated soundness of the parts of the pedestrian bridge; and outputting the decided course of action.

[0104] Furthermore, some or all of the configurations described in Supplements 2 to 18, which are dependent on Supplement 1 described above, may also be dependent on Supplements 19 and 20 in the same dependent relationship as Supplements 2 to 18. Furthermore, not limited to Supplement 1, Supplement 19, and Supplement 20, some or all of the configurations described as Supplements may be made dependent on various hardware, software, various recording means for recording software, or systems, within the scope of each of the above-mentioned embodiments.

[0105] Although the present disclosure has been described above with reference to the embodiments, the present disclosure is not limited to the above-described embodiments. Various modifications that can be understood by those skilled in the art can be made to the configuration and details of the present disclosure within the scope of the present disclosure. Furthermore, each embodiment can be combined with other embodiments as appropriate.

[0106] REFERENCE SIGNS LIST 10 Footbridge diagnosis system 11 Acquisition unit 12 Detection unit 13 Estimation unit 14 Determination unit 15 Prediction unit 16 Output unit 17 Storage unit 20 On-board device 30 Terminal device 100 Computer 101 CPU 102 Memory 103 Storage device 104 Input / output I / F 105 Communication I / F

Claims

1. A pedestrian bridge diagnostic system comprising: an acquisition means for acquiring images of a pedestrian bridge installed on a road, taken using a device mounted on a mobile body; a detection means for detecting parts of the pedestrian bridge that appear in the acquired images; an estimation means for estimating the soundness of the parts of the pedestrian bridge based on the detected images of the parts of the pedestrian bridge; a decision means for determining a course of action for repairing the pedestrian bridge based on the estimated soundness of the parts of the pedestrian bridge; and an output means for outputting the decided course of action.

2. The pedestrian bridge diagnostic system according to claim 1, wherein the estimation means determines a course of action for repairing the pedestrian bridge based on the soundness of multiple parts of the pedestrian bridge.

3. A pedestrian bridge diagnostic system as described in claim 1 or 2, wherein the detection means detects parts of the pedestrian bridge in each of images taken of the same pedestrian bridge from at least two directions, and the estimation means estimates the soundness of each part of the pedestrian bridge based on the images of the parts of the pedestrian bridge taken from at least two directions.

4. A pedestrian bridge diagnostic system according to claim 3, wherein the detection means detects the portion of the pedestrian bridge in images taken from each of the roads that intersect at the location where the pedestrian bridge is installed.

5. A pedestrian bridge diagnostic system according to any one of claims 1 to 4, wherein the estimation means estimates the soundness of parts of the pedestrian bridge based on a reference image and an image acquired by the acquisition means.

6. The pedestrian bridge diagnostic system according to claim 5, wherein the reference image is an image taken when the pedestrian bridge has been repaired or installed.

7. A pedestrian bridge diagnostic system as described in any one of claims 1 to 6, wherein the determination means determines a course of action for repairing the pedestrian bridge based on the soundness of parts of the pedestrian bridge and the environment in which the pedestrian bridge is installed.

8. A pedestrian bridge diagnostic system as described in any one of claims 1 to 7, wherein the estimation means estimates the soundness of the part of the pedestrian bridge based on an image of the part of the pedestrian bridge and the environment in which the pedestrian bridge is installed.

9. A pedestrian bridge diagnostic system according to any one of claims 1 to 8, wherein the determining means determines a course of action for repairing the pedestrian bridge further based on the importance of the pedestrian bridge.

10. A pedestrian bridge diagnostic system as claimed in any one of claims 1 to 9, further comprising a prediction means for predicting when the pedestrian bridge will need to be repaired or removed, and wherein the output means further outputs the prediction result of when the pedestrian bridge will need to be repaired or removed.

11. A pedestrian bridge diagnosis system according to any one of claims 1 to 10, wherein the output means outputs the response policy for each of the pedestrian bridges by superimposing it on a map.

12. A pedestrian bridge diagnostic system according to any one of claims 1 to 11, wherein the output means outputs the soundness of each of the pedestrian bridges by superimposing it on a map.

13. A pedestrian bridge diagnostic system according to any one of claims 1 to 12, wherein the estimation means estimates the soundness further based on the shape of the pedestrian bridge.

14. A pedestrian bridge diagnostic system as described in any one of claims 1 to 13, wherein the estimation means estimates the healthiness of each part of the pedestrian bridge that appears in the image acquired by the acquisition means using an estimation model that estimates the healthiness of parts of the pedestrian bridge from the image in which the parts appear.

15. A pedestrian bridge diagnostic system according to claim 14, wherein the estimation means estimates the soundness of the part of the pedestrian bridge using an estimation model based on the part of the pedestrian bridge.

16. A pedestrian bridge diagnostic system according to any one of claims 1 to 15, wherein the detection means identifies the pedestrian bridge based on characters written on the pedestrian bridge.

17. A pedestrian bridge diagnostic system as described in any one of claims 1 to 16, wherein the estimation means estimates the healthiness of parts of the pedestrian bridge that are not shown in the image based on the healthiness of multiple parts of the pedestrian bridge.

18. A pedestrian bridge diagnostic system according to any one of claims 1 to 17, wherein the images are time-series video images taken using a device mounted on a vehicle.

19. A pedestrian bridge diagnosis method comprising: acquiring an image of a pedestrian bridge installed on a road photographed using a device mounted on a mobile body; detecting the part of the pedestrian bridge that appears in the acquired image; estimating the soundness of the part of the pedestrian bridge based on the detected image of the part of the pedestrian bridge; determining a response policy for the pedestrian bridge based on the estimated soundness of the part of the pedestrian bridge; and outputting the determined response policy.

20. A recording medium that non-temporarily records a pedestrian bridge diagnosis program that causes a computer to execute the following processes: acquiring an image of a pedestrian bridge installed on a road, taken using a device mounted on a mobile body; detecting the parts of the pedestrian bridge that appear in the acquired image; estimating the soundness of the parts of the pedestrian bridge based on the detected image of the parts of the pedestrian bridge; determining a course of action for repairing the pedestrian bridge based on the estimated soundness of the parts of the pedestrian bridge; and outputting the determined course of action.

Images