WO2020261568A1 - 路面検査装置、路面検査方法、およびプログラム - Google Patents

路面検査装置、路面検査方法、およびプログラム Download PDF

Info

Publication number
WO2020261568A1
WO2020261568A1 PCT/JP2019/025950 JP2019025950W WO2020261568A1 WO 2020261568 A1 WO2020261568 A1 WO 2020261568A1 JP 2019025950 W JP2019025950 W JP 2019025950W WO 2020261568 A1 WO2020261568 A1 WO 2020261568A1
Authority
WO
WIPO (PCT)
Prior art keywords
road
damage
road surface
judgment
surface inspection
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/JP2019/025950
Other languages
English (en)
French (fr)
Japanese (ja)
Inventor
健一 山岬
中野　学
慎一郎 鷲見
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
NEC Corp
Original Assignee
NEC Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by NEC Corp filed Critical NEC Corp
Priority to PCT/JP2019/025950 priority Critical patent/WO2020261568A1/ja
Priority to JP2021527305A priority patent/JP7156527B2/ja
Priority to US17/620,180 priority patent/US20220254169A1/en
Publication of WO2020261568A1 publication Critical patent/WO2020261568A1/ja
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/50—Context or environment of the image
    • G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
    • G06V20/588—Recognition of the road, e.g. of lane markings; Recognition of the vehicle driving pattern in relation to the road
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0004—Industrial image inspection
    • E—FIXED CONSTRUCTIONS
    • E01—CONSTRUCTION OF ROADS, RAILWAYS, OR BRIDGES
    • E01C—CONSTRUCTION OF, OR SURFACES FOR, ROADS, SPORTS GROUNDS, OR THE LIKE; MACHINES OR AUXILIARY TOOLS FOR CONSTRUCTION OR REPAIR
    • E01C23/00—Auxiliary devices or arrangements for constructing, repairing, reconditioning, or taking-up road or like surfaces
    • E01C23/01—Devices or auxiliary means for setting-out or checking the configuration of new surfacing, e.g. templates, screed or reference line supports; Applications of apparatus for measuring, indicating, or recording the surface configuration of existing surfacing, e.g. profilographs
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
    • G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/50—Context or environment of the image
    • G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/70—Labelling scene content, e.g. deriving syntactic or semantic representations
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30108—Industrial image inspection
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30181—Earth observation
    • G06T2207/30184—Infrastructure
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30248—Vehicle exterior or interior
    • G06T2207/30252—Vehicle exterior; Vicinity of vehicle
    • G06T2207/30256—Lane; Road marking

Definitions

  • the present invention relates to a technique for supporting the management work of the laid road.
  • Patent Document 1 An example of a technique for efficiently inspecting a road is disclosed in Patent Document 1 below.
  • Patent Document 1 below discloses an example of a technique for detecting damage (cracks, ruts, etc.) on the road surface using an image of a road.
  • the accuracy is high. At present, the accuracy is improved by visually confirming the judgment result of road damage by a computer. However, it is very troublesome to visually confirm all the determination results of all the images. A technology that improves the accuracy of road damage judgment results by a computer while reducing the work load by humans is desired.
  • the present invention has been made in view of the above problems.
  • One of the objects of the present invention is to provide a technique for improving the accuracy of the judgment result of road damage by a computer while reducing the work load by a person.
  • the first road surface inspection device of the present invention is Image acquisition means to acquire the input image showing the road, A damage detecting means for detecting a damaged part of a road in the input image by using a damage determining device for determining a damaged part of the road constructed by machine learning. An output means that outputs a judgment result having a certainty level of less than or equal to a reference value among the judgment results of the damaged part of the road by the damage judgment device to the display device in a state that can be distinguished from other judgment results. To be equipped.
  • the second road surface inspection device of the present invention is Image acquisition means to acquire the input image showing the road, A damage detecting means for detecting a damaged part of a road in the input image by using a damage determining device for determining a damaged part of the road constructed by machine learning. An output means for outputting the judgment result of the damaged part of the road by the damage judgment device to the display device together with the certainty of the judgment result. To be equipped.
  • the first road surface inspection method of the present invention is The computer Get the input image of the road and Using a damage determiner that determines the damaged part of the road constructed by machine learning, the damaged part of the road in the input image is detected. Among the judgment results of the damaged part of the road by the damage judgment device, the judgment result whose certainty is equal to or less than the reference value is output to the display device in a state where it can be distinguished from other judgment results. Including that.
  • the second road surface inspection method of the present invention is The computer Get the input image of the road and Using a damage determiner that determines the damaged part of the road constructed by machine learning, the damaged part of the road in the input image is detected.
  • the judgment result of the damaged part of the road by the damage judgment device is output to the display device together with the certainty of the judgment result. Including that.
  • the program of the present invention causes a computer to execute the above-mentioned first road surface inspection method or second road surface inspection method.
  • FIG. 1st Embodiment It is a figure which illustrates the functional structure of the road surface inspection apparatus in 1st Embodiment. It is a block diagram which illustrates the hardware composition of the road surface inspection apparatus. It is a flowchart which illustrates the flow of the process executed by the road surface inspection apparatus of 1st Embodiment. It is a figure which shows an example of the screen which is output to the display device by an output part. It is a figure which shows another example of the screen which is output to the display device by an output part. It is a figure which shows another example of the screen which is output to the display device by an output part. It is a figure which illustrates the functional structure of the road surface inspection apparatus in 2nd Embodiment.
  • each block diagram represents a configuration of a functional unit, not a configuration of a hardware unit.
  • the direction of the arrow in the figure is for making the flow of information easy to understand, and does not limit the direction of communication (one-way communication / two-way communication) unless otherwise specified.
  • FIG. 1 is a diagram illustrating a functional configuration of a road surface inspection device according to the first embodiment.
  • the road surface inspection device 10 illustrated in FIG. 1 includes an image acquisition unit 110, a damage detection unit 120, and an output unit 130.
  • the image acquisition unit 110 acquires an input image showing the road surface to be inspected. As illustrated in FIG. 1, the image of the road surface is generated by the image pickup device 22 provided in the vehicle 20. Specifically, while the vehicle 20 is traveling on the road in the inspection target section, the image pickup device 22 performs a photographing operation to generate a road surface image of the road in the inspection target section.
  • the image acquisition unit 110 acquires at least one of a plurality of frame images constituting the road surface image as an image to be image-processed (analyzed).
  • the image acquisition unit 110 may acquire an image of the road surface from the image pickup device 22 via the network.
  • the image acquisition unit 110 may acquire a road surface image from, for example, an image pickup device 22 connected by a communication cable or a portable storage medium such as a memory card.
  • the damage detection unit 120 detects a damaged part of the road in the input image acquired by the image acquisition unit 110 by using the damage determination device 122.
  • the damage determination device 122 selects the damaged part of the road from the input image by repeating machine learning using the learning data in which the image of the road and the information indicating the damaged part of the road (correct answer label) are combined. It is constructed so that it can be judged.
  • the learning data used when the damage determination device 122 is first constructed is generated, for example, by a data analyst performing an operation of assigning an appropriate correct answer label to the learning image.
  • the damage determiner 122 is modeled by machine learning to detect, for example, cracks, ruts, potholes, depressions, depressions, and steps on the road surface as damaged parts of the road.
  • the output unit 130 outputs the determination result of the damaged portion of the road by the damage determination device 122 to the display device 30.
  • the output unit 130 sets a determination result having a certainty level equal to or lower than the reference value among the determination results of the damaged portion of the road by the damage determination device 122 as another determination result (a determination result having a certainty degree exceeding the reference value). It is output to the display device 30 in a distinguishable state.
  • the degree of certainty is information indicating the certainty of the damage determination result by the damage determination device 122.
  • confidence is represented by a binary value of 0 (low confidence) and 1 (high confidence), or a continuous value in the range 0 to 1.
  • the damage determining device 122 uses, for example, the degree of similarity between the feature amount of the damaged part of the road obtained by machine learning and the feature amount obtained from the damaged part (pixel area) reflected in the input image as the certainty of the determination result. Can be calculated.
  • Each functional component of the road surface inspection device 10 may be realized by hardware that realizes each functional component (eg, a hard-wired electronic circuit, etc.), or a combination of hardware and software (eg, example). It may be realized by a combination of an electronic circuit and a program that controls it).
  • FIG. 2 is a block diagram illustrating a hardware configuration of the road surface inspection device 10.
  • the road surface inspection device 10 includes a bus 1010, a processor 1020, a memory 1030, a storage device 1040, an input / output interface 1050, and a network interface 1060.
  • the bus 1010 is a data transmission path for the processor 1020, the memory 1030, the storage device 1040, the input / output interface 1050, and the network interface 1060 to transmit and receive data to and from each other.
  • the method of connecting the processors 1020 and the like to each other is not limited to the bus connection.
  • the processor 1020 is a processor realized by a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), or the like.
  • the memory 1030 is a main storage device realized by a RAM (Random Access Memory) or the like.
  • the storage device 1040 is an auxiliary storage device realized by an HDD (Hard Disk Drive), an SSD (Solid State Drive), a memory card, a ROM (Read Only Memory), or the like.
  • the storage device 1040 stores a program module that realizes each function of the road surface inspection device 10 (image acquisition unit 110, damage detection unit 120, output unit 130, etc.).
  • the processor 1020 reads each of these program modules into the memory 1030 and executes them, each function corresponding to each program module is realized.
  • the input / output interface 1050 is an interface for connecting the road surface inspection device 10 and various input / output devices.
  • An input device such as a keyboard or mouse, an output device (not shown) such as a display or a printer may be connected to the input / output interface 1050.
  • the input / output interface 1050 may be connected to an image pickup device 22 (or a portable storage medium provided in the image pickup device 22) or a display device 30.
  • the road surface inspection device 10 can communicate with the image pickup device 22 (or a portable storage medium provided in the image pickup device 22) via the input / output interface 1050 to acquire the road surface image generated by the image pickup device 22. .. Further, the road surface inspection device 10 can output the screen generated by the output unit 130 to the display device 30 connected via the input / output interface 1050.
  • the network interface 1060 is an interface for connecting the road surface inspection device 10 to the network.
  • This network is, for example, LAN (Local Area Network) or WAN (Wide Area Network).
  • the method of connecting the network interface 1060 to the network may be a wireless connection or a wired connection.
  • the road surface inspection device 10 can communicate with the image pickup device 22 and a video database (not shown) via the network interface 1060 to acquire the road surface video generated by the image pickup device 22. Further, the road surface inspection device 10 can communicate with the display device 30 via the network interface 1060 and display the screen generated by the output unit 130 on the display device 30.
  • the hardware configuration of the road surface inspection device 10 is not limited to the configuration illustrated in FIG.
  • FIG. 3 is a flowchart illustrating a flow of processing executed by the road surface inspection device 10 of the first embodiment.
  • the image acquisition unit 110 acquires an input image (an image of the road to be processed) (S102).
  • the image acquisition unit 110 acquires the road surface image generated by the image pickup apparatus 22 via, for example, the input / output interface 1050 and the network interface 1060.
  • the image acquisition unit 110 reads all or a part of the plurality of frame images constituting the road surface image as an image of the road to be processed.
  • the image acquisition unit 110 may be configured to execute pre-processing on the image of the road in order to improve the efficiency of processing in the post-process.
  • the image acquisition unit 110 may execute preprocessing such as front correction processing and deblurring processing on the road image.
  • the damage detection unit 120 detects the damaged part of the road from the input image using the damage determination device 122 (S104).
  • the damage detection unit 120 acquires information indicating the position (damaged portion of the road) determined to be road damage in the input image and information indicating the degree of certainty regarding the determination from the damage determination device 122.
  • the damage determination device 122 determines a pixel region having a certain degree of similarity with the feature amount of the damaged portion of the road obtained by machine learning as the damaged portion of the road, and outputs the determination result.
  • the damage determination device 122 determines the similarity between the feature amount of the damaged part of the road obtained by machine learning and the feature amount calculated from the feature amount extracted from the pixel region determined as the damaged part of the road. , Output as the certainty of the judgment result.
  • the damage detection unit 120 acquires this information as "a determination result of a damaged portion of the road by the damage determination device 122".
  • the output unit 130 outputs the determination result of the damaged portion of the road by the damage determination device 122 (S106).
  • the output unit 130 determines whether or not the determination result of the damaged portion of the road by the damage determination device 122 includes the determination result having the certainty degree of the reference value or less.
  • the output unit 130 compares the certainty of each judgment result of the damaged part of the road by the damage judgment device 122 with a preset reference value, so that the certainty is less than or equal to the reference value (specifically, , The pixel area corresponding to the determination result whose certainty is less than or equal to the reference value) is specified.
  • the output unit 130 can distinguish the determination result from other determination results.
  • the state is output to the display device 30.
  • the screen output to the display device 30 by the output unit 130 will be illustrated below.
  • FIG. 4 is a diagram showing an example of a screen output to the display device 30 by the output unit 130.
  • the output unit 130 determines that the certainty is equal to or less than the reference value and other determination results whose certainty exceeds the reference value, depending on the display mode of the specific display element (rectangular frame). Are distinguishable from each other. Specifically, the output unit 130 assigns a rectangular frame A shown by a solid line to a portion determined to be "road damage” with a certainty level exceeding the reference value. Further, the output unit 130 assigns a rectangular frame B shown by a dotted line to a portion determined to be “road damage” or “not road damage” with a certainty level equal to or lower than the reference value.
  • the output unit 130 does not add a specific display element such as a rectangular frame to the portion determined to be "not damage to the road” with a certainty level equal to or higher than the reference value.
  • a specific display element such as a rectangular frame
  • the result of judgment with low certainty that is, the judgment result to be confirmed by human eyes
  • the output unit 130 further outputs the determination result of whether or not the road is a damaged portion and the character information C indicating the certainty of the determination result.
  • the output unit 130 may be configured to include information indicating the type of road damage (cracks, potholes, etc.) in the character information C. ..
  • FIG. 20 is a diagram showing another example of the screen output to the display device 30 by the output unit 130.
  • the output unit 130 is configured to switch the color of the frame line, the thickness of the frame line, the fill pattern in the frame, and the like based on whether or not the certainty level corresponding to the determination result is equal to or less than the reference value. You may. Further, for example, the output unit 130 may be configured to set the color of the frame line, the thickness of the frame line, the fill pattern in the frame, and the like according to the certainty level corresponding to the determination result. Further, the output unit 130 may use a display element other than the rectangular frame as a display element to be given to each of the determination results by the damage determination device 122.
  • the output unit 130 emphasizes the shape of the damaged part of the road (the shape of a crack or a pothole) in order to make it possible to distinguish between the judgment result whose certainty is less than the reference value and other judgment results (the shape of the crack or pothole).
  • a line or fill that emphasizes the outer shape may be used.
  • the output unit 130 is a display element that emphasizes the shape of an object that is not determined to be a "damaged portion of the road" in a certain area as a result of the determination by the damage determination device 122.
  • a line or fill that emphasizes the outer shape may be output.
  • the output unit 130 is configured so that the determination result with low certainty can be distinguished from other determination results by a specific display element such as a rectangular frame without displaying the character information C. It may be good (example: FIG. 5).
  • FIG. 5 is a diagram showing another example of the screen output to the display device 30 by the output unit 130. Even with the screen illustrated in FIG. 5, the determination result with low certainty can be easily identified by the display mode (solid line / dotted line) of the rectangular frame.
  • the output unit 130 may change the display mode of a specific display element based on the determination result (damage / non-damage determination) of the damaged portion of the road and the certainty of the determination result (example: FIG. 6).
  • FIG. 6 is a diagram showing another example of the screen output to the display device 30 by the output unit 130.
  • the output unit 130 assigns a rectangular frame A shown by a solid line to a portion determined to be "road damage” with a certainty exceeding the reference value.
  • the output unit 130 assigns a rectangular frame B shown by a dotted line to a portion determined to be "damaged to the road” with a certainty level equal to or lower than the reference value.
  • the output unit 130 assigns a rectangular frame D indicated by a dotted line and filled with diagonal lines to a portion determined to be "not a road damage” with a certainty level equal to or lower than the reference value. It should be noted that the output unit 130 does not add a specific display element such as a rectangular frame to the portion determined to be “not damage to the road” with a certainty level equal to or higher than the reference value. According to the screen illustrated in FIG. 6, the part judged as "road damage” with the certainty of less than the standard value (that is, the part with a relatively high possibility of detection error) and the certainty of less than the standard value. The part determined to be “not road damage” (that is, the part with a relatively high possibility of omission of detection) can be further identified.
  • the determination result that the certainty is equal to or less than the reference value that is, the certainty is low.
  • "low certainty" of the judgment result of the damaged part of the road by the damage judgment device 122 means that the judgment result includes an error (erroneous detection or omission of detection) when viewed by the human eye. It can be said that the possibility is relatively high.
  • the work of visually confirming the presence or absence of a determination error by the damage determination device 122 is streamlined, and the workload is reduced as a whole. The effect of making it can be expected.
  • the output unit 130 may be configured to output the determination result of the damaged portion of the road by the damage determination device 122 together with the certainty of the determination result.
  • the output unit 130 is configured to output a display element (character information C) indicating the certainty of the determination result by the damage determination device 122 for each determination result of the damaged portion of the road. Will be done.
  • FIG. 21 is a diagram showing another example of the screen output to the display device 30 by the output unit 130. As illustrated in FIG. 21, by visualizing the magnitude of the certainty of the judgment result by the damage judgment device 122, the judgment result having a high possibility of error (that is, the judgment result that should be confirmed with particular care). Can be easily grasped by a person viewing the screen output to the display device 30.
  • the road surface inspection device 10 of the present embodiment is different from the above-described first embodiment in that it further has a configuration related to the correction work as described below.
  • FIG. 7 is a diagram illustrating the functional configuration of the road surface inspection device 10 according to the second embodiment.
  • the road surface inspection device 10 of the present embodiment further includes a damage determination result correction unit 140 and a first learning unit 150.
  • the damage determination result correction unit 140 corrects the determination result that is the target of the correction input based on the correction input for the determination result of the damaged part of the road output to the display device 30. Specifically, a person performing the confirmation work of the screen output on the display device 30 (the screen displaying the judgment result of the damaged part of the road by the damage judgment device 122) makes an erroneous judgment found on the screen. An input operation (correction input) for correcting the result to a correct determination result is performed using the input device 40. The damage determination result correction unit 140 receives this correction input via the input device 40. Then, the damage determination result correction unit 140 corrects the erroneous determination result based on the correction input.
  • the damage determination result correction unit 140 determines that the operation is the target of the operation. Correct the result and update the display contents on the screen (reflect the correction).
  • the first learning unit 150 generates teacher data (first teacher data) for machine learning of the damage determination device 122 by using the correction input and the input image for the determination result of the damaged portion. For example, the first learning unit 150 extracts a partial image area corresponding to the determination result that is the target of the correction input, and indicates the partial image area and the determination result (indicating the damaged part / non-damaged part of the road) indicated by the correction input.
  • the first teacher data can be generated in combination with the correct answer label). Further, the first learning unit 150 may generate the first teacher data by combining the input image acquired by the image acquisition unit 110 and the determination result of the damaged portion of the road by the damage determination device 122.
  • the judgment result of the damaged part of the road by the damage judgment device 122 includes the judgment result corrected by the damage judgment result correction unit 140 as the target of the correction input and the judgment result not to be the target of the correction input. obtain. Then, the first learning unit 150 learns (re-learns) the damage determining device 122 using the generated first teacher data.
  • FIG. 8 is a flowchart illustrating a flow of processing executed by the road surface inspection device 10 of the second embodiment. The processing shown below is executed after the output processing by the output unit 130 (example: the processing of S106 in FIG. 3).
  • the damage determination result correction unit 140 receives a correction input for the determination result of the damaged part of the road by the damage determination device 122 (S202). This correction input is performed by a person who confirms the screen displayed on the display device 30 by using an input device 40 such as a keyboard, a mouse, and a touch panel. Then, the damage determination result correction unit 140 corrects the determination result that is the target of the correction input based on the correction input (S204).
  • FIG. 9 A specific example of the operation by the damage determination result correction unit 140 will be described with reference to the figure. 9, FIG. 10, FIG. 22 and FIG. 23 are diagrams illustrating a specific operation of the damage determination result correction unit 140. It should be noted that these are merely examples, and the operation of the damage determination result correction unit 140 is not limited to the contents disclosed in these figures.
  • correction information for the determination result is input via the user interface E as illustrated in FIGS. 9 and 22.
  • an input is made to correct the determination of "road damage” by the damage determining device 122 as “not road damage (non-damaged)".
  • the damage determination result correction unit 140 corrects the target determination result based on the input in the user interface E.
  • the display on the display device 30 is updated as shown in FIG. 10, for example. In the screen illustrated in FIG.
  • the damage determination result correction unit 140 since the part targeted for the correction input was designated by a person as "a part that is not a damage to the road" (determined with a high degree of certainty), the damage determination result correction unit 140 , The rectangular frame that was displayed for that part is hidden. Further, in the example of FIG. 22, an input is made to correct the determination of "not damage to the road (non-damaged)" by the damage determination device 122 to "damage to the road".
  • the user operates the pointer P using an input device or the like, and the damaged part of the road that the damage judge 122 could not detect (“It is not a road damage” by the damage judge 122 ( The judgment result) "non-damaged)" is specified.
  • the damage determination result correction unit 140 corrects the target determination result based on the input in the user interface E.
  • the display on the display device 30 is updated as shown in FIG. 23, for example.
  • the damage determination result correction unit 140 is the part.
  • the screen display is updated so that the rectangular frame displayed for is shown by a solid line.
  • the first learning unit 150 generates the first teacher data using the correction input received in S202 and the input image acquired by the image acquisition unit 110 (S206). For example, the first learning unit 150 extracts a partial image corresponding to the determination result to be corrected by the correction input from the input image, the image feature amount of the partial image or the partial image, and the content of the correction input. Combined with (information indicating road damage / non-damage), the first teacher data is generated. Then, the first learning unit 150 executes the learning process of the damage determining device 122 using the generated first teacher data (S208). The first learning unit 150 may be configured to execute the learning process of the damage determining device 122 each time the correction input is received.
  • the first learning unit 150 stores the first teacher data generated in response to the correction input in a predetermined storage area, and performs a learning process using the stored first teacher data at a predetermined timing. It may be configured to be executed (for example, at the timing of regular night maintenance).
  • Example of action / effect As described above, according to the present embodiment, when there is an error in the determination result of the damaged portion of the road by the damage determination device 122, it is possible to correct the error by human judgment. Further, in the present embodiment, teacher data for machine learning of the damage determination device 122 is generated according to a correction input for the determination result of the damaged portion of the road by the damage determination device 122, and the damage determination device is used using the teacher data. 122 relearning processes are executed. As a result, it is possible to improve the judgment system of the damaged portion of the road by the damage judgment device 122 and reduce the number of appearances of the judgment result with low certainty (the judgment result that a person should confirm).
  • the work of correcting the determination error of the damaged portion of the road by the damage determination device 122 also serves as the work of generating teacher data for machine learning. Therefore, even if the conventional work of generating the learning data (the work of manually associating the learning image data with the correct label) is not performed separately, the damage determination device 122 is included in the output confirmation work by the output unit 130. Can generate training data for. As a result, the labor required to improve the accuracy of the damage determining device 122 can be reduced.
  • FIG. 11 is a diagram illustrating the functional configuration of the road surface inspection device 10 according to the third embodiment.
  • a plurality of divisions are defined in the width direction of the road.
  • the plurality of divisions include, for example, roadways, shoulders, sidewalks, and the ground adjacent to the road (the area outside the road adjacent to the road).
  • the damage detection unit 120 includes a plurality of damage determination devices 122 corresponding to each of the plurality of categories as described above.
  • Each damage determination device 122 is constructed by machine learning so as to determine a damaged portion in each of a plurality of sections set in the width direction of the road.
  • a certain damage determination device 122 repeats machine learning by repeating machine learning using teacher data that combines an image for learning and information (correct answer label) indicating the position of a damaged part of the roadway in the image. It is built as a judgment device specialized for judging damaged parts of the road. Further, another damage determination device 122 repeats machine learning by using teacher data that combines an image for learning and information (correct answer label) indicating the position of a damaged part of the sidewalk in the image. , It is built as a judgment device specialized in judging the damaged part of the sidewalk. Further, machine learning is also performed on the division of the ground adjacent to the road shoulder or the road, and a damage determination device 122 specialized for determining the damaged portion of each division is constructed.
  • the damage detection unit 120 of the present embodiment uses a plurality of damage determination devices 122 to detect a damaged portion of the road for each road division as described above.
  • the damage detection unit 120 has a division determination device 124 for determining an area corresponding to each of a plurality of divisions defined in the width direction of the road.
  • the damage detection unit 120 of the present embodiment uses the classification determination device 124 to specify a region corresponding to each of the above-mentioned plurality of divisions in the input image acquired by the image acquisition unit 110.
  • the classification determination device 124 is defined in the width direction of the road by repeating machine learning using learning data that combines an image of the road and information (correct answer label) indicating the classification of the road reflected in the image.
  • the area corresponding to each of the plurality of divisions can be determined from the image.
  • the output unit 130 outputs the determination results of the above-mentioned plurality of divisions by the division determination device 124 to the display device 30 together with the determination results of the damaged portion of the road by the damage determination device 122.
  • the road surface inspection device 10 of the present embodiment may further include the damage determination result correction unit 140 and the first learning unit 150 described in the second embodiment.
  • FIG. 12 is a flowchart illustrating a flow of processing executed by the road surface inspection device 10 of the third embodiment.
  • the image acquisition unit 110 acquires the input image to be processed (S302). This is the same as the process of S102 in the flowchart of FIG.
  • the damage detection unit 120 identifies an image area corresponding to each division of the road from the input images acquired by the image acquisition unit 110 by using the classification determination device 124 (S304). Then, the damage detection unit 120 detects the damaged portion of the road for each division from the input image acquired by the image acquisition unit 110 by using the damage determination device 122 corresponding to the division specified in the process of S304. (S306). At this time, the damage detection unit 120 identifies the image area of the division corresponding to each of the plurality of damage determination devices 122 from the input image by using the determination result of each division by the classification determination device 124, and a plurality of the specified image areas. It may be input to each of the damage determination devices 122 of. By doing so, it is possible to improve the accuracy of the output (determination result of the damaged portion of the road) of each of the plurality of damage determination devices 122.
  • the output unit 130 determines the road classification determination result by the classification determination device 124 obtained by the processing of S304 and the damage portion of the road for each division by the damage determination device 122 for each division obtained by the processing of S306.
  • the result is output to the display device 30 (S308).
  • the output unit 130 outputs, for example, a screen as illustrated in FIG. 13 to the display device 30.
  • FIG. 13 is a diagram showing an example of a screen output to the display device 30 by the output unit 130 of the third embodiment.
  • the output unit 130 displays the judgment result of the road classification by the classification judgment device 124 in addition to the display elements A to C showing the judgment result of the road damage portion by the plurality of damage judgment devices 122.
  • the display elements F1 to F3 shown are further output.
  • the determination result of the road division by the division determination device 124 is further output via the display device 30.
  • a person can visually recognize how the machine (road surface inspection device 10) recognizes the road reflected in the input image. Further, based on the judgment result of the road classification and the judgment result of the damaged part of the road by the damage judgment device 122 constructed for each road classification, how each damage judgment device 122 determines the damaged part of the road. A person can visually recognize whether or not the judgment has been made.
  • the plurality of damage assessors 122 are classified by road surface materials such as "asphalt” and “concrete” instead of (or in addition to) the widthwise division of the road such as “road” and “sidewalk”. You may.
  • the classification determination device 124 is constructed so that the road surface material of the road shown in the image can be identified instead of (or in addition to) the classification such as a roadway or a sidewalk.
  • the classification determination device 124 repeats machine learning using teacher data that combines an image of a road for learning and a correct label indicating the road surface material of the image, thereby performing a feature amount for each road surface material of the road. You can learn.
  • the damage detection unit 120 acquires information indicating the road surface material of the road shown in the image to be processed from the classification determination device 124, selects the damage determination device 122 corresponding to the road surface material indicated by the information, and determines the damage of the road surface. Determine the presence or absence. According to the configuration of this modification, the optimum learning model (damage judge 122) is selected according to the road surface material of the road shown in the image to be processed, so that the effect of improving the detection accuracy of the road surface damage is expected. it can.
  • the present embodiment has the same configuration as the first embodiment, the second embodiment, or the third embodiment described above, except for the points described below.
  • FIG. 14 is a diagram illustrating the functional configuration of the road surface inspection device 10 according to the fourth embodiment. As illustrated in FIG. 14, the road surface inspection device 10 of the present embodiment further includes a classification determination result correction unit 160 and a second learning unit 170.
  • the classification determination result correction unit 160 determines the classification of the road that is the target of the classification correction input based on the correction input for the road classification determination result by the classification determination device 124 (hereinafter, also referred to as “classification correction input”). Correct the judgment result. Specifically, a person who confirms the screen output on the display device 30 (the screen displaying the road classification judgment result by the classification judgment device 124 and the judgment result of the damaged part of the road by the damage judgment device 122).
  • the input device 40 is used to perform an input operation (classification correction input) for correcting an erroneous judgment result regarding the road classification found on the screen to a correct judgment result.
  • the classification determination result correction unit 160 receives the classification correction input via the input device 40. Then, the classification determination result correction unit 160 corrects an erroneous determination result regarding the road classification based on the classification correction input.
  • the second learning unit 170 generates teacher data (second teacher data) for machine learning of the classification judgment device 124 by using the classification correction input and the input image for the judgment result by the classification judgment device 124. For example, the second learning unit 170 extracts a partial image area corresponding to the determination result that is the target of the division correction input, and the determination result indicated by the partial image area and the division correction input (correct answer indicating the type of road division). The second teacher data can be generated in combination with the label). Further, the second learning unit 170 may generate the first teacher data by combining the input image acquired by the image acquisition unit 110 and the determination result of the road division by the division determination device 124.
  • the judgment result of the road classification by the classification judgment device 124 is the judgment result corrected by the classification judgment result correction unit 160 as the target of the classification correction input and the judgment result not to be the target of the classification correction input.
  • the second learning unit 170 learns (re-learns) the classification determination device 124 using the generated second teacher data.
  • FIG. 15 is a flowchart illustrating a flow of processing executed by the road surface inspection device 10 of the fourth embodiment. The processing shown below is executed after the output processing by the output unit 130 (example: the processing of S106 in FIG. 3).
  • the classification determination result correction unit 160 accepts the classification correction input for the road classification determination result by the classification determination device 124 (S402). This division correction input is performed by a person who confirms the screen displayed on the display device 30 by using an input device 40 such as a keyboard, a mouse, and a touch panel. Then, the damage determination result correction unit 140 corrects the determination result that is the target of the division correction input based on the division correction input (S404).
  • 16 and 17 are diagrams illustrating a specific operation of the classification determination result correction unit 160. It should be noted that these are merely examples, and the operation of the classification determination result correction unit 160 is not limited to the contents disclosed in these figures.
  • correction information for the judgment result of the classification is input by the operation as illustrated in FIG.
  • the classification determination result correction unit 160 relates to the determination result regarding the classification of the "roadway” (the area determined to be the "roadway” in the image) and the classification of the "sidewalk” as shown in FIG. Correct the judgment result (the area judged as "roadway” in the image).
  • the classification determination result correction unit 160 can perform, for example, an input operation for deforming a part of the shape or boundary line of each division or an input operation for resetting the shape or boundary of the division. It may be configured to provide a user interface.
  • the second learning unit 170 generates the second teacher data using the correction input received in S402 and the input image acquired by the image acquisition unit 110 (S406). For example, the second learning unit 170 extracts a partial image area corresponding to the determination result to be corrected by the division correction input from the input image, and determines the partial image area or the image feature amount of the partial image area.
  • the second teacher data is generated by combining with the content of the classification correction input (information indicating the type of road classification). Then, the second learning unit 170 executes the learning process of the classification determination device 124 using the generated second teacher data (S408).
  • the second learning unit 170 may be configured to execute the learning process of the classification determination device 124 each time the classification correction input is received.
  • the second learning unit 170 stores the second teacher data generated in response to the division correction input in a predetermined storage area, and performs a learning process using the stored second teacher data. It may be configured to be executed at a timing (for example, the timing of regular night maintenance).
  • the teacher data for the division determination device 124 is generated according to the division correction input received by the division determination result correction unit 160, and the re-learning of the division determination device 124 is executed.
  • the determination accuracy of the road division of the division determination device 124 is improved, and an appropriate input can be given to the plurality of damage determination devices 122 specially constructed for each of the plurality of divisions.
  • the detection accuracy of the damaged portion of the road for each category is improved, and the number of occurrences of judgment results with low certainty (judgment results to be confirmed by a person) can be reduced. By reducing the number of occurrences of judgment results with low certainty, further efficiency of the entire work can be expected.
  • the work of correcting the road classification determination error by the classification determination device 124 also serves as the work of generating teacher data for machine learning. Therefore, even if the conventional work of generating the learning data (the work of manually associating the learning image data with the correct label) is not performed separately, the classification determination device 124 is included in the output confirmation work by the output unit 130. Can generate training data for. As a result, the labor required to improve the accuracy of the classification determination device 124 and the damage determination device 122 can be reduced.
  • the road surface inspection device 10 of the present embodiment has a function of generating teacher data of the damage determination device 122 and executing machine learning by using the road determination result of the damage determination device 122. Different from.
  • FIG. 18 is a block diagram illustrating the functional configuration of the road surface inspection device 10 of the fifth embodiment. As shown in FIG. 18, the road surface inspection device 10 of the present embodiment further includes an image acquisition unit 110, a damage detection unit 120, and a learning unit 180.
  • the damage detection unit 120 includes a plurality of damage determination devices 122 for determining a damaged portion of a road, and a classification determination device 124 for determining a road classification.
  • each of the plurality of damage determination devices 122 includes a plurality of pre-defined divisions for the road (width divisions such as "roadway”, “shoulder”, and “sidewalk”, and road surfaces such as “asphalt” and “concrete”. It corresponds to the classification of materials, etc.).
  • the learning unit 180 generates teacher data used for machine learning of the damage determining device 122 by using the determination result of the damaged part of the road by the damage determining device 122 and the input image. Then, the learning unit 180 executes machine learning of the damage determining device 122 using the generated teacher data.
  • the learning unit 180 is configured to select the damage determining device 122 to be the target of machine learning using the generated teacher data based on the classification determination result by the classification determining device 124.
  • the learning unit 180 has obtained information indicating that the road surface material of the road is "asphalt" as a result of the classification determination by the classification determination device 124.
  • the learning unit 180 selects the damage determining device 122 corresponding to the "asphalt" category as the target of machine learning using the generated teacher data.
  • the learning unit 180 has obtained information indicating that the division in the width direction of the road is a "roadway” as a determination result of the division by the division determination device 124.
  • the learning unit 180 selects the damage determining device 122 corresponding to the classification of the "roadway” as the target of machine learning using the generated teacher data. Further, the learning unit 180 has obtained information indicating that the road surface material of the road is "asphalt” and the division in the width direction of the road is "roadway” as a result of the classification determination by the classification determination device 124. To do. In this case, the learning unit 180 selects the damage determining device 122 corresponding to the classification of "asphalt" and "roadway” as the target of machine learning using the generated teacher data. By doing so, the possibility that the damage determination device 122 learns an erroneous feature amount by the teacher data (noise data) of different divisions is reduced. As a result, it is possible to prevent the determination accuracy of the damage determination device 122 from being lowered by machine learning.
  • the damaged part of the road may be located over two or more sections.
  • road cracks may extend from the roadway to the shoulder.
  • the learning unit 180 is configured to select the damage determining device 122 to be machine-learned based on the magnitude (number of pixels) of the road damage in each of the two or more divisions. Good.
  • the learning unit 180 uses the teacher data generated by using the image including the cracked portion of the road.
  • the damage determination device 122 for the roadway is selected as the target of machine learning using the above.
  • the learning unit 180 may be configured to generate teacher data for each of the two or more sections using the damaged portion of the road in each of the two or more sections. For example, as shown in FIG. 19, when the damage to the road is located over the two sections of the road and the shoulder, the learning unit 180 covers the image area indicated by the reference numeral G (the area indicated by the broken line in the figure). It is configured to generate the teacher data of the damage judge 122 for the road by using it, and to generate the teacher data of the damage judge 122 for the road shoulder using the image area indicated by the symbol H (the area indicated by the dotted line in the figure). You may be.
  • Image acquisition means to acquire the input image showing the road
  • a damage detecting means for detecting a damaged part of a road in the input image by using a damage determining device for determining a damaged part of the road constructed by machine learning.
  • An output means that outputs a judgment result having a certainty level of less than or equal to a reference value among the judgment results of the damaged part of the road by the damage judgment device to the display device in a state that can be distinguished from other judgment results.
  • a road surface inspection device equipped with 2. 2.
  • Image acquisition means to acquire the input image showing the road
  • a damage detecting means for detecting a damaged part of a road in the input image by using a damage determining device for determining a damaged part of the road constructed by machine learning.
  • An output means for outputting the judgment result of the damaged part of the road by the damage judgment device to the display device together with the certainty of the judgment result.
  • a damage determination result correction means for correcting the determination result that is the target of the correction input is further provided based on the correction input for the determination result of the damaged portion of the road output to the display device. 1. 1. Or 2. The road surface inspection device described in. 4.
  • a first learning means for generating first teacher data using the correction input and the input image and learning the damage determining device using the first teacher data is further provided.
  • the damage detecting means detects a damaged portion of a road in each of the plurality of divisions by using the damage determining device constructed for each of the plurality of divisions. 1. 1. From 4. The road surface inspection device according to any one of the above. 6. The damage detecting means identifies a region corresponding to each of the plurality of divisions in the input image by using a division determining device for determining a region corresponding to each of the plurality of divisions constructed by machine learning. The output means further outputs the determination results of the plurality of divisions by the division determination device to the display device. 5.
  • a classification determination result correction means for correcting the determination result that is the target of the division correction input is further provided based on the division correction input for the determination results of the plurality of categories output to the display device. 6.
  • the road surface inspection device described in. 8 A second learning means for generating second teacher data using the division correction input and the input image and learning the division determination device using the second teacher data is further provided. 7.
  • the road surface inspection device described in. 9. The computer Get the input image of the road and Using a damage determiner that determines the damaged part of the road constructed by machine learning, the damaged part of the road in the input image is detected.
  • the judgment result whose certainty is equal to or less than the reference value is output to the display device in a state where it can be distinguished from other judgment results.
  • Road surface inspection method including that. 10.
  • the computer Get the input image of the road and Using a damage determiner that determines the damaged part of the road constructed by machine learning, the damaged part of the road in the input image is detected.
  • the judgment result of the damaged part of the road by the damage judgment device is output to the display device together with the certainty of the judgment result.
  • Road surface inspection method including that. 11.
  • the computer Based on the correction input for the judgment result of the damaged part of the road output to the display device, the judgment result that is the target of the correction input is corrected. Including that 9. Or 10.
  • the computer The first teacher data is generated by using the correction input and the input image, and the damage determination device is learned by using the first teacher data. Including 11. Road surface inspection method described in. 13. Multiple divisions are defined for roads, The computer The damaged portion of the road is detected for each of the plurality of divisions by using the damage determining device constructed for each of the plurality of divisions. Including that 9. From 12. The road surface inspection method according to any one of the above. 14. The computer Using a division determination device that determines the region corresponding to each of the plurality of divisions constructed by machine learning, the region corresponding to each of the plurality of divisions is specified in the input image. The determination results of the plurality of divisions by the classification determination device are further output to the display device. Including 13.
  • Road surface inspection method described in. 15. The computer Based on the classification correction input for the judgment results of the plurality of classifications output to the display device, the judgment result that is the target of the classification correction input is corrected. Including 14. Road surface inspection method described in. 16.
  • the computer The second teacher data is generated by using the division correction input and the input image, and the classification determination device is learned by using the second teacher data. Including 15. Road surface inspection method described in. 17. To the computer, 9. From 16. A program for executing the road surface inspection method described in any one of the above. 18.
  • Image acquisition means to acquire the input image showing the road A damage detecting means for detecting a damaged part of a road from the input image by using a damage determining device for determining a damaged part of the road constructed by machine learning.
  • the damage determiner is constructed for each of a plurality of road-related sections.
  • the damage detecting means is Based on the input image, the category corresponding to the road reflected in the image is determined from the plurality of categories.
  • the damaged part of the road is detected by using the damage judge corresponding to the judged classification.
  • the learning means The damage determination device to be learned is selected based on the determination result of the classification by the damage detection means.
  • the learning means When the damaged part of the road is located over two or more sections out of the plurality of sections, the learning is performed based on the size of the damaged part of the road in each of the two or more sections. Select a damage judge, 18. The road surface inspection device described in. 20. The learning means When the damaged part of the road is located over two or more of the plurality of sections, the damaged part of the road in each of the two or more sections is used, and the damage corresponding to each of the two or more sections is used. Generate teacher data for the judge, 18. The road surface inspection device described in.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Databases & Information Systems (AREA)
  • Medical Informatics (AREA)
  • Health & Medical Sciences (AREA)
  • Computing Systems (AREA)
  • Computational Linguistics (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Software Systems (AREA)
  • Quality & Reliability (AREA)
  • Architecture (AREA)
  • Civil Engineering (AREA)
  • Structural Engineering (AREA)
  • Image Analysis (AREA)
  • Road Repair (AREA)
  • Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
PCT/JP2019/025950 2019-06-28 2019-06-28 路面検査装置、路面検査方法、およびプログラム Ceased WO2020261568A1 (ja)

Priority Applications (3)

Application Number Priority Date Filing Date Title
PCT/JP2019/025950 WO2020261568A1 (ja) 2019-06-28 2019-06-28 路面検査装置、路面検査方法、およびプログラム
JP2021527305A JP7156527B2 (ja) 2019-06-28 2019-06-28 路面検査装置、路面検査方法、およびプログラム
US17/620,180 US20220254169A1 (en) 2019-06-28 2019-06-28 Road surface inspection apparatus, road surface inspection method, and program

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/JP2019/025950 WO2020261568A1 (ja) 2019-06-28 2019-06-28 路面検査装置、路面検査方法、およびプログラム

Publications (1)

Publication Number Publication Date
WO2020261568A1 true WO2020261568A1 (ja) 2020-12-30

Family

ID=74061563

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2019/025950 Ceased WO2020261568A1 (ja) 2019-06-28 2019-06-28 路面検査装置、路面検査方法、およびプログラム

Country Status (3)

Country Link
US (1) US20220254169A1 (https=)
JP (1) JP7156527B2 (https=)
WO (1) WO2020261568A1 (https=)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP7229432B1 (ja) * 2022-04-08 2023-02-27 三菱電機株式会社 施設管理情報表示装置、施設管理情報表示システム、施設管理情報表示方法および施設管理情報表示プログラム
JP2023035618A (ja) * 2021-09-01 2023-03-13 ダイハツ工業株式会社 異常検知装置および異常検知方法
JP7391117B2 (ja) 2022-01-07 2023-12-04 三菱電機株式会社 車両用画像処理装置および車両用画像処理方法

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20240176185A (ko) * 2023-06-15 2024-12-24 현대자동차주식회사 노면의 가혹도 추정 장치

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0142807B2 (https=) * 1984-02-04 1989-09-14 Kodama Chemical Ind
US20170204569A1 (en) * 2016-01-15 2017-07-20 Fugro Roadware Inc. High speed stereoscopic pavement surface scanning system and method
JP2017167969A (ja) * 2016-03-17 2017-09-21 首都高技術株式会社 損傷抽出システム
JP2018154979A (ja) * 2017-03-16 2018-10-04 株式会社リコー 情報処理システム、評価システム、情報処理方法およびプログラム
JP2019056668A (ja) * 2017-09-22 2019-04-11 エヌ・ティ・ティ・コムウェア株式会社 情報処理装置、情報処理システム、情報処理方法、及び情報処理プログラム

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2017014288A1 (ja) * 2015-07-21 2017-01-26 株式会社東芝 ひび割れ解析装置、ひび割れ解析方法及びひび割れ解析プログラム
CN106548182B (zh) * 2016-11-02 2020-02-18 武汉理工大学 基于深度学习和主成因分析的路面裂纹检测方法及装置
US11104345B2 (en) * 2018-04-18 2021-08-31 Rivian Ip Holdings, Llc Methods, systems, and media for determining characteristics of roads
JP6442807B1 (ja) * 2018-06-15 2018-12-26 カラクリ株式会社 対話サーバ、対話方法及び対話プログラム
CN109190481B (zh) * 2018-08-06 2021-11-23 中国交通通信信息中心 一种遥感影像道路材质提取方法和系统

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0142807B2 (https=) * 1984-02-04 1989-09-14 Kodama Chemical Ind
US20170204569A1 (en) * 2016-01-15 2017-07-20 Fugro Roadware Inc. High speed stereoscopic pavement surface scanning system and method
JP2017167969A (ja) * 2016-03-17 2017-09-21 首都高技術株式会社 損傷抽出システム
JP2018154979A (ja) * 2017-03-16 2018-10-04 株式会社リコー 情報処理システム、評価システム、情報処理方法およびプログラム
JP2019056668A (ja) * 2017-09-22 2019-04-11 エヌ・ティ・ティ・コムウェア株式会社 情報処理装置、情報処理システム、情報処理方法、及び情報処理プログラム

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2023035618A (ja) * 2021-09-01 2023-03-13 ダイハツ工業株式会社 異常検知装置および異常検知方法
JP7535484B2 (ja) 2021-09-01 2024-08-16 ダイハツ工業株式会社 異常検知装置および異常検知方法
JP7391117B2 (ja) 2022-01-07 2023-12-04 三菱電機株式会社 車両用画像処理装置および車両用画像処理方法
JP7229432B1 (ja) * 2022-04-08 2023-02-27 三菱電機株式会社 施設管理情報表示装置、施設管理情報表示システム、施設管理情報表示方法および施設管理情報表示プログラム

Also Published As

Publication number Publication date
JPWO2020261568A1 (https=) 2020-12-30
US20220254169A1 (en) 2022-08-11
JP7156527B2 (ja) 2022-10-19

Similar Documents

Publication Publication Date Title
JP7517489B2 (ja) 路面検査装置、路面検査方法、およびプログラム
WO2020261568A1 (ja) 路面検査装置、路面検査方法、およびプログラム
Radopoulou et al. Automated detection of multiple pavement defects
Zhou et al. Crack segmentation through deep convolutional neural networks and heterogeneous image fusion
US10824906B2 (en) Image processing device, non-transitory computer readable storage medium, and image processing system
CN110443783B (zh) 一种图像质量评估方法及装置
US11164031B2 (en) System and method for analyzing an image of a vehicle
KR20090101356A (ko) 결함 검출 장치 및 결함 검출 방법
US20220317055A1 (en) Inspection apparatus, control method, and program
CN104103257A (zh) 数据处理装置和包括数据处理装置的显示系统
JP6980208B2 (ja) 構造物維持管理業務支援システム
JP2021165888A (ja) 情報処理装置、情報処理装置の情報処理方法およびプログラム
CN119723448B (zh) 一种基于时序监测数据的施工道路裂缝预测方法
CN101826159B (zh) 实现灰度图像分区二值化的方法和数据处理设备
JP6894361B2 (ja) コンクリート表面上のひび割れ方向特定方法、ひび割れ方向特定装置、ひび割れ方向特定システム及びプログラム
CN119130912A (zh) 基于深度图的带钢表面缺陷检测方法及系统
CN110824451A (zh) 雷达回波图的处理方法、装置、计算机设备及存储介质
CN118230271A (zh) 车道线检测方法、模型训练方法、装置、设备及存储介质
CN115690090B (zh) 缺陷样本图像生成方法、装置、设备及存储介质
CN114998889B (zh) 一种沉浸式三维影像的智能识别方法及系统
CN110263789B (zh) 一种物体边界识别方法、装置及设备
CN117788444A (zh) Smt贴片偏移检测方法、装置及视觉检测系统
CN113033891B (zh) 高墩桥梁健康监控方法及装置
WO2022097275A1 (ja) 情報処理システム
CN115115680A (zh) 图像处理方法、装置、设备以及存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19935434

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2021527305

Country of ref document: JP

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19935434

Country of ref document: EP

Kind code of ref document: A1