Reinforcement inspection device, reinforcement inspection method, and program

Abstract

To provide a bar arrangement inspection device, a bar arrangement inspection method and a program which can specify a measurement range of a reinforcing iron bar from an inspection region and distinguish the reinforcing iron bar being the measurement object.SOLUTION: A bar arrangement inspection device 1 comprises: an identifier detection unit 11 which acquires from an imaging unit an image obtained by imaging an inspection region in which a plurality of reinforcing iron bars is arranged and a tape in which a plurality of markers is provided with an interval in a longitudinal direction and which is arranged below the reinforcing iron bar being the measurement object in the inspection region and detects the marker from the acquired image; an identifier position information calculation unit 12 which calculates identifier position information indicating the position of the detected marker; a measurement range specification unit 13 which specifies a measurement range of the reinforcing iron bar using the identifier position information; a photographed image acquisition unit 14 which acquires photographed image data of the measurement range from the imaging unit 2; a bar arrangement detection unit 15 which detects reinforcing iron bar position information indicating the position of the reinforcing iron bar over the tape by using the photographed image data of the measurement range; and a measurement processing unit 16 which measures the reinforcing iron bar corresponding to the reinforcing iron bar position information.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

[0002] In the construction of reinforced concrete structures, after the reinforcing bars have been assembled, a reinforcing bar arrangement inspection is performed to check whether the reinforcing bars have been arranged as designed. For example, Patent Document 1 describes a reinforcing bar arrangement inspection system that uses image data of an inspection area where reinforcing bars have been arranged to measure the number of reinforcing bars, reinforcing bar spacing, reinforcing bar diameter, etc. in the inspection area, and then performs a reinforcing bar arrangement inspection using these measurement results.

[0003] In the reinforcement inspection system described in Patent Document 1, if the image obtained from the first capture does not include the entire inspection area, additional images are taken of the remaining part of the inspection area that did not fit into the image, the measurement range is identified from the image, and measurement processing is performed. This series of processes is repeated until measurement of the entire inspection area is completed. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Publication No. 2020-166566 Summary of the Invention [Problem to be solved by the invention]

[0005] In the conventional technology described in Patent Document 1, when the inspection area is wide, the inspector needs to visually identify the measurement range. However, since the inspection area generally has reinforcing bars arranged periodically (for example, in a grid pattern), it is difficult for the inspector to visually distinguish between partial areas of the inspection area, making it difficult to sequentially identify the measurement range from the inspection area.

[0006] Furthermore, the inspection area viewed three-dimensionally is a structure in which multiple planes, each with multiple reinforcing bars, are stacked. In conventional technology, it was necessary to analyze the three-dimensional structure of the structure using three-dimensional image data of the inspection area in order to identify the plane on which the reinforcing bars to be measured are placed from among the multiple stacked planes.

[0007] The present disclosure is intended to solve the above-mentioned problems, and aims to provide a reinforcement inspection device, a reinforcement inspection method, and a program that can identify the rebar measurement range from the inspection area without relying on the inspector's visual inspection, and can identify the rebar to be measured without analyzing the three-dimensional structure of the inspection area. [Means for solving the problem]

[0008] The reinforcement inspection device according to the present disclosure includes an identifier detection unit that acquires from a photographing unit an image of an inspection area in which a plurality of reinforcing bars are arranged, and a strip-shaped member having a plurality of identifiers spaced apart in the longitudinal direction and positioned beneath the reinforcing bars to be measured in the inspection area, and detects the identifier from the acquired image; an identifier position information calculation unit that calculates identifier position information indicating the position of the identifier; a measurement range identification unit that uses the identifier position information to identify the measurement range of the reinforcing bars; an image acquisition unit that acquires photographed image data of the measurement range from the photographing unit; a reinforcement detection unit that uses the photographed image data of the measurement range to detect reinforcing bar position information indicating the position of the reinforcing bars spanning over the strip-shaped member; and a measurement processing unit that measures the reinforcing bars corresponding to the reinforcing bar position information. [Effects of the Invention]

[0009] According to the present disclosure, an inspection device detects identifiers from a captured image of an inspection area in which multiple rebars are arranged and a strip-shaped member having multiple identifiers spaced apart in the longitudinal direction and positioned in the inspection area beneath the rebar to be measured, calculates identifier position information indicating the position of the detected identifier, identifies a measurement range of the rebar using the identifier position information, detects rebar position information indicating the position of the rebar straddling the strip-shaped member using captured image data of the measurement range, and measures the rebar corresponding to the rebar position information. The rebar inspection device according to the present disclosure automatically identifies the measurement range of the rebar using the identifier position information, so that the measurement range of the rebar can be identified from the inspection area without visual inspection by an inspector. In addition, the rebar inspection device according to the present disclosure identifies the rebar straddling the strip-shaped member as the measurement target, so that the rebar to be measured can be identified without analyzing the three-dimensional structure of the inspection area. [Brief explanation of the drawings]

[0010] [Figure 1] 1 is a block diagram showing the configuration of a bar arrangement inspection device according to a first embodiment. [Figure 2] FIG. 10 is a screen view showing an image of an inspection area in which tape has been laid. [Figure 3] 3 is a flowchart showing a reinforcement bar arrangement inspection method according to the first embodiment. [Figure 4] FIG. 10 is an explanatory diagram showing an outline of a photographing process of an inspection area. [Figure 5] FIG. 10 is a front view showing a tape on which a marker is written. [Figure 6] FIG. 10 is a screen diagram showing an example of a display of an image of an inspection area before reaching a photographing position. [Figure 7] FIG. 10 is a screen diagram showing an example of a display of an image of an inspection area when the imaging position is reached. [Figure 8] 10 is a flowchart showing details of a reinforcing bar measurement process. [Figure 9] FIG. 10 is an explanatory diagram showing a process of converting a captured image into a normal image. [Figure 10] FIG. 2 is a block diagram showing the configuration of a measurement unit and a learning device. [Figure 11]FIG. 10 is an explanatory diagram showing a process of extracting a partial image of a reinforcing bar from a normalized image. [Figure 12] FIG. 10 is an explanatory diagram showing the process of inputting a partial image of a reinforcing bar into a learning model to infer the diameter of the reinforcing bar. [Figure 13] 13A and 13B are block diagrams showing a hardware configuration for realizing the functions of the bar arrangement inspection device according to the first embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0011] Embodiment 1 1 is a block diagram showing the configuration of a reinforcing bar inspection device 1 according to embodiment 1. The reinforcing bar inspection device 1 specifies a measurement range for measuring reinforcing bars from an inspection area in which multiple reinforcing bars are arranged, and measures inspection information for the reinforcing bars using captured image data for the specified measurement range. The inspection information for the reinforcing bars includes, for example, the number of reinforcing bars arranged in the inspection area (number of arranged reinforcing bars), the spacing between the reinforcing bars (reinforcing bar spacing), the diameter of the reinforcing bars, and the spacing between nodes.

[0012] The bar arrangement inspection device 1 is connected wirelessly or via a wire to the imaging unit 2 and the display unit 3. For example, the bar arrangement inspection device 1 is a smartphone, tablet terminal, or laptop (Personal Computer) equipped with the imaging unit 2 and the display unit 3. In the following explanation, it is assumed that the bar arrangement inspection device 1 is a tablet terminal.

[0013] The photographing unit 2 is a camera that photographs the inspection area. For example, a monocular camera can be used as the photographing unit 2. A monocular camera is a camera that is cheaper than a stereo camera, and as mentioned above, it may be a camera attached to a smartphone or tablet terminal. However, the photographing unit 2 may also be a stereo camera. A stereo camera is composed of left and right cameras arranged at a certain distance from each other. Furthermore, the photographing unit 2 may be one of the left and right cameras that make up the stereo camera.

[0014] The display unit 3 displays the image captured by the imaging unit 2, the captured image data, and the imaging conditions or the measurement results of the rebars in the inspection area. For example, the display unit 3 is a display provided on a smartphone or tablet terminal, and is a liquid crystal display or organic EL display.

[0015] The video displayed on the display unit 3 is time-series data of still images (frame images) captured by the photographing unit 2 and acquired at a fixed cycle by the reinforcing bar inspection device 1. The fixed cycle is assumed to be, for example, 30 (fps) at most. The captured image data is image data acquired by the reinforcing bar inspection device 1 as still images of the measurement range where reinforcing bar measurements are performed by operating the photographing unit 2.

[0016] A touch panel is mounted on the screen of the display unit 3. When an image for photographing operations is displayed on the display unit 3, the photographing unit 2 starts photographing an image when an operation on the displayed image is received using the touch panel.

[0017] FIG. 2 is a screen diagram showing an image 2A of the inspection area 4 where a tape 6 has been laid. The image 2A is captured by the imaging unit 2 and displayed on the display unit 3. The tape 6 is a strip-shaped member with a plurality of markers 7 provided at intervals along its length. For example, the plurality of markers 7 are provided at equal intervals along the length of the tape 6. Furthermore, the plurality of markers 7 are identifiers to which unique identification information (e.g., an identification number) is assigned.

[0018] When the inspection area 4 is viewed three-dimensionally, it has a structure in which multiple planes, each with multiple reinforcing bars 5 arranged therein, are stacked, i.e., it has an architectural or civil engineering structure with multiple reinforcing bars 5 arranged as a skeleton. For this reason, in addition to the reinforcing bars 5 on the plane closest to the photographing unit 2, image 2A of the inspection area 4 also reflects reinforcing bars 5 on planes farther from the photographing unit 2 than this plane, as shown by the dashed line in Figure 2.

[0019] In conventional technology, in order to exclude rebars 5 on a plane different from the plane of interest from the measurement target, the three-dimensional structure of the inspection area 4 is analyzed using three-dimensional image data of the inspection area 4, and only the rebars 5 on the plane of interest are extracted from the three-dimensional structure of the inspection area 4. Since analyzing the three-dimensional structure of the inspection area 4 imposes a computational load, if this analysis is not necessary, the computational load can be significantly reduced. Furthermore, in order to obtain three-dimensional image data of the inspection area 4, it has generally been necessary to use an expensive camera such as a stereo camera (three-dimensional camera, three-dimensional laser scanner).

[0020] The tape 6 is placed under the rebar 5 to be measured in the inspection area 4. The inspector moves the monocular camera along the tape 6 laid in the inspection area 4. The reinforcing bar inspection device 1 uses the image data of the inspection area 4 captured by the monocular camera to identify the measurement range of the rebar 5 and the rebar 5 to be measured. As shown in Figure 2, the image 2A of the inspection area captured by the monocular camera captures the rebar 5 and the tape 6.

[0021] Because the tape 6 is placed under the rebar 5 to be measured, the rebar arrangement inspection device 1 can identify the rebar 5 that straddles the tape 6 in the image 2A as the rebar 5 to be measured. This allows the rebar arrangement inspection device 1 to determine the inspection information of the rebar 5 to be measured without analyzing the three-dimensional structure of the inspection area 4, eliminating the need for an expensive computing device or an expensive stereo camera.

[0022] 1, the reinforcement bar inspection device 1 includes an identifier detection unit 11, an identifier position information calculation unit 12, a measurement range specification unit 13, a captured image acquisition unit 14, a reinforcement bar detection unit 15, a measurement processing unit 16, and a display processing unit 17. The identifier detection unit 11 acquires, from the imaging unit 2, an image of a tape 6 placed under a rebar 5 to be measured, and detects a marker 7 from the acquired image.

[0023] The identifier position information calculation unit 12 calculates identifier position information indicating the position of the detected marker 7. For example, the identifier position information calculation unit 12 calculates the position of the marker 7 on the frame image constituting the video of the inspection area. The identifier position information is the two-dimensional coordinate position of the marker 7 on the frame image constituting the video.

[0024] The measurement range specifying unit 13 specifies the measurement range of the rebar 5 from the inspection area using the identifier position information. For example, the measurement range specifying unit 13 uses the identifier position information to specify, as the measurement range, the range defined by the lines connecting adjacent markers 7. The measurement range specifying unit 13 outputs the position information of the specified measurement range to the captured image acquisition unit 14.

[0025] The captured image acquisition unit 14 acquires captured image data of the measurement range from the imaging unit 2. For example, the captured image acquisition unit 14 acquires captured image data of the measurement range captured by a monocular camera, which is the imaging unit 2. For example, when the measurement range is identified from the image displayed on the display unit 3, the captured image acquisition unit 14 outputs an instruction signal to the imaging unit 2 and the display processing unit 17 to permit the imaging unit 2 to capture the inspection area.

[0026] When the photographing unit 2 receives an instruction signal from the photographed image acquisition unit 14, it enters a state of waiting for a photographing operation. When the display processing unit 17 receives an instruction signal from the photographed image acquisition unit 14, it displays an image for photographing operation on the display unit 3. When the image for photographing operation is operated using the touch panel, the photographing unit 2 starts photographing. The photographed image acquisition unit 14 acquires photographed image data photographed by the photographing unit 2.

[0027] The photographing unit 2 may automatically photograph the measurement range when it receives an instruction signal from the photographed image acquiring unit 14. The photographed image acquiring unit 14 acquires the photographed image data automatically photographed by the photographing unit 2. In this case, the inspector does not need to perform a photographing operation each time the measurement range is identified by the measurement range identifying unit 13. Therefore, the reinforcement arrangement inspection device 1 can sequentially acquire photographed image data of the measurement range simply by moving the photographing unit 2 along the tape 6 laid in the inspection area.

[0028] The reinforcing bar detection unit 15 uses the captured image data of the measurement range to detect reinforcing bar position information indicating the positions of the reinforcing bars 5 spanning the tape 6. For example, the reinforcing bar detection unit 15 performs image analysis such as edge analysis on the captured image of the measurement range to detect images of the reinforcing bars 5 in the captured image of the measurement range, and calculates the two-dimensional coordinate position information of the reinforcing bars 5 on the captured image as the reinforcing bar position information. The reinforcing bar position information may be position information of each point in a group of points that form the axis of the reinforcing bar 5 on the photographed image, or may be a function that approximates the axis of the reinforcing bar 5.

[0029] The measurement processing unit 16 measures the rebars 5 corresponding to the rebar position information. For example, the measurement processing unit 16 measures the number of rebars 5, rebar spacing, rebar diameter, and node spacing within each measurement range. Next, the measurement processing unit 16 outputs to the display processing unit 17 the measurement results, including the sum of the number of rebars for each measurement range, the average rebar spacing for each measurement range, the diameter of each rebar 5 in all measurement ranges, and the node spacing of the rebars 5 in all measurement ranges.

[0030] The display processing unit 17 displays information on the display unit 3. For example, the display processing unit 17 inputs an image from the photographing unit 2 via the identifier detection unit 11 and displays the input image on the display unit 3. The display processing unit 17 also displays the rebar measurement results output from the measurement processing unit 16 on the display unit 3.

[0031] The display processing unit 17 may be provided separately from the reinforcement inspection device 1 and may be a component included in a display device having the display unit 3. In this case, the reinforcement inspection device 1 will include the identifier detection unit 11, identifier position information calculation unit 12, measurement range specification unit 13, photographed image acquisition unit 14, reinforcement detection unit 15, and measurement processing unit 16, excluding the display processing unit 17. Furthermore, when the reinforcement inspection device 1 displays the measurement results of the reinforcing bars 5 on the display unit 3, the reinforcement inspection device 1 accesses the display device and transmits display information showing the measurement results from the reinforcement inspection device 1 to the display device. The display device displays the display information received from the reinforcement inspection device 1 on the display unit 3.

[0032] The measurement processing unit 16 also includes an image conversion unit 161, an image extraction unit 162, and a measurement unit 163. The image conversion unit 161 converts the captured image of the measurement range indicated by the captured image data acquired by the captured image acquisition unit 14 into a oriented image. The oriented image is an image in which the distance between the photographing unit 2 and a plane including the rebar 5 to be measured in the inspection area is scaled to a constant value, and the plane is converted so that it appears as if it is facing the photographing unit 2 directly.

[0033] For example, the image conversion unit 161 specifies the four corner points of any rectangle on the reinforcing bars 5 arranged in a grid pattern on the plane to be measured, and estimates a homography transformation matrix using the position coordinates of the specified four points. Then, the image conversion unit 161 converts the captured image into an orthogonalized image based on the estimated homography transformation matrix. All points included in the orthogonalized image are scaled so that their distance from the image capture unit 2 is constant. Therefore, in the orthogonalized image, differences in the size of the reinforcing bars 5 depending on their distance from the image capture unit 2 are corrected. The image conversion unit 161 corrects the reinforcing bar position information calculated by the reinforcing bar arrangement detection unit 15 to match it with the orientated image, and outputs the corrected reinforcing bar position information to the image extraction unit 162.

[0034] The image extraction unit 162 extracts an image of the reinforcing bar 5 from the normal image based on the reinforcing bar position information indicating the position of the reinforcing bar 5 in the measurement range, and outputs the extracted image to the measurement unit 163. Furthermore, the image extraction unit 162 may extract multiple partial images from the image of the reinforcing bar 5 and output them to the measurement unit 163. The multiple partial images are obtained by dividing the partial image of the reinforcing bar 5 of the same image size in order along the longitudinal direction of the reinforcing bar 5.

[0035] The measurement unit 163 measures the reinforcing bars 5 using the images of the reinforcing bars 5 extracted from the orientated image by the image extraction unit 162. For example, the measurement unit 163 measures the number of reinforcing bars 5 included in the measurement range (number of reinforcing bars) and the interval between adjacent reinforcing bars 5 (reinforcing bar interval) based on the images of the reinforcing bars 5 and the reinforcing bar position information. The measurement unit 163 measures the diameter of the reinforcing bars 5 and the interval between nodes by performing image analysis on the images of the reinforcing bars 5.

[0036] In addition, the measurement unit 163 may infer at least one of the diameter and the spacing of the reinforcing bars 5 using a learning model that takes multiple partial images extracted by the image extraction unit 162 as input and outputs at least one of the diameter and the spacing of the reinforcing bars 5 shown in the partial images.

[0037] 3 is a flowchart showing the reinforcement bar inspection method according to the first embodiment, illustrating a series of processes performed by the reinforcement bar inspection device 1. The reinforcement bar inspection device 1 is a tablet terminal having a photographing unit 2 and a display unit 3. The inspector moves along the longitudinal direction of the tape 6 laid in the inspection area while pointing the photographing direction of the photographing unit 2 toward the inspection area.

[0038] The identifier detection unit 11 acquires the image of the inspection area 4 captured by the imaging unit 2 as it moves along the tape 6, and detects the marker 7 from the image (step ST1). In this way, the imaging unit 2 sequentially captures the entire inspection area 4. For example, the identifier detection unit 11 detects the marker 7 by performing image analysis such as pattern matching on the image of the inspection area 4.

[0039] Fig. 4 is an explanatory diagram showing an overview of the photographing process of the inspection area 4. Fig. 5 is a front view showing a tape 6 on which a marker 7 is written. As shown in Fig. 4, the inspector lays the tape 6 parallel to the rebar 5 at the end of the inspection area 4 and under the rebar 5 to be measured. As shown in Fig. 5, the marker 7 is an identifier in which an identification shape 7B is drawn on a square base 7A, and is, for example, an AR (augmented reality) marker.

[0040] As shown in FIG. 4 , when an inspector moves the bar arrangement inspection device 1 along the tape 6, the photographing unit 2 photographs an image of the inspection area 4 including the tape 6. The identifier detection unit 11 acquires an image, which is time-series data of frame images, from the image photographed by the photographing unit 2, for example, at a cycle of 30 fps. The identifier detection unit 11 outputs the image photographed by the photographing unit 2 to the display processing unit 17. The display processing unit 17 displays the image output from the identifier detection unit 11 on the display unit 3.

[0041] 3, when the identifier detection unit 11 detects the marker 7, the identifier position information calculation unit 12 calculates identifier position information indicating the position of the detected marker 7 (step ST2). For example, the identifier position information calculation unit 12 binarizes the frame image to detect the contour of the background 7A of the marker 7, and calculates the identifier position information when it is determined that the background 7A is a square based on the detected contour.

[0042] 5, when the shape formed by the line segment connecting vertex P1 and vertex P2, the line segment connecting vertex P2 and vertex P3, the line segment connecting vertex P3 and vertex P4, and the line segment connecting vertex P4 and vertex P1 is a square, the frame image corresponds to an image captured from the front of the inspection area 4. The capturing position in this case is a position where the imaging unit 2 is located between adjacent markers 7.

[0043] If any of the four vertices P1 to P4 of the base 7A is not detected or the shape of the base 7A is detected distorted, it is considered that the frame image at that time was captured by the imaging unit 2 from an oblique direction of the inspection area 4. An image in which the inspection area is captured from an oblique direction cannot accurately identify adjacent measurement ranges in the inspection area 4. For this reason, the identifier position information calculation unit 12 identifies the marker 7 in the frame image captured at a position where the base 7A is captured as a square, and outputs the identifier position information of the identified marker 7 to the measurement range identification unit 13.

[0044] The measurement range specifying unit 13 uses the identifier position information to specify the measurement range 8 in which to measure the reinforcing bars (step ST3). For example, as shown in Fig. 4, when the reinforcement bar arrangement inspection device 1 is positioned between adjacent markers 7, that is, when two adjacent markers 7 are captured in the frame image, the measurement range specifying unit 13 determines the range defined by the line connecting the adjacent markers 7 to be the measurement range 8.

[0045] FIG. 6 is a screen diagram showing an example of a display of an image of the inspection area 4 before the device reaches the photographing position. In FIG. 6, a sub-screen (hereinafter referred to as a window screen) M1 of the image of the inspection area 4 is displayed on the display screen 3A of the display unit 3. The window screen M1 displays an image that is orientated directly relative to the photographing unit 2. However, in reality, as will be described later with reference to FIG. 9, an image is displayed in which the size of the reinforcing bar 5 on the image changes depending on the distance between the photographing unit 2 and the plane of the measurement target. As shown in FIG. 6, only one marker 7 is displayed on the window screen M1. This means that the reinforcement bar arrangement inspection device 1 has not yet moved to the photographing position where two adjacent markers 7 are displayed, as shown in FIG. 4. The display processing unit 17 displays the peripheral portion N1 of the window screen M1, for example, in red on the display screen 3A.

[0046] When the window screen M1 with the red border N1 is displayed on the display screen 3A, the imaging unit 2 is not permitted to capture an image of the inspection area 4. In this case, the display processing unit 17 either processes the imaging operation image 2B into a semi-transparent image indicating that imaging operation is not permitted and displays it on the display screen 3A, as shown by the dashed line in FIG. 6, or does not display the imaging operation image 2B. In this case, the examiner cannot operate the imaging operation image 2B, and the imaging unit 2 only captures an image of the inspection area 4 but cannot capture an image of the measurement range. This prevents an image of the inspection area 4 from being erroneously captured when the measurement range has not been specified.

[0047] FIG. 7 is a screen diagram showing an example of the image of the inspection area 4 displayed when the camera reaches the shooting position. In FIG. 7, a window screen M2 of the image of the inspection area 4 is displayed on the display screen 3B of the display unit 3. The window screen M2 displays an image that is orientated directly relative to the camera 2. However, in reality, as will be described later with reference to FIG. 9, the image displayed changes in size depending on the distance between the camera 2 and the plane of the measurement target. As shown in FIG. 7, two markers 7 are displayed on the window screen M2. This means that the reinforcement bar arrangement inspection device 1 is located at the shooting position where two adjacent markers 7 are displayed, as shown in FIG. 4. At this time, the display processing unit 17 displays the peripheral portion N2 of the window screen M2, for example, in green on the display screen 3B.

[0048] When the window screen M2 with the green border N2 is displayed on the display screen 3B, the photographing unit 2 is permitted to photograph the inspection area 4. The display processing unit 17 displays the photographing operation image 2B on the display screen 3B. The examiner performs a touch operation on the photographing operation image 2B, causing the photographing unit 2 to photograph an image of the measurement range.

[0049] The display processing unit 17 may display the image of the inspection area 4 displayed on the display unit 3 in different modes before the measurement range is identified and when the measurement range is identified. For example, instead of changing the color of the periphery of the window screen displaying the image, the shape of the window screen may be changed. Furthermore, the measurement range identifying unit 13 may instruct an audio output device (not shown in FIG. 1) to notify by audio that photography is permitted when the reinforcement bar arrangement inspection device 1 moves between two adjacent markers 7.

[0050] 3, when the measurement range is specified, the captured image acquisition unit 14 acquires captured image data of the measurement range captured by the imaging unit 2 (step ST4). For example, the captured image acquisition unit 14 acquires captured image data of the measurement range captured by the imaging unit 2 in response to a touch operation on the image 2A by the examiner. Note that if the imaging unit 2 is a monocular camera, the captured image data is two-dimensional image data of the measurement range.

[0051] The reinforcing bar arrangement detection unit 15 detects reinforcing bar position information indicating the positions of the reinforcing bars 5 in the measurement range using the photographed image data acquired by the photographed image acquisition unit 14 (step ST5). The inspection area 4 generally has not one layer but multiple layers of planes on which multiple reinforcing bars 5 are arranged in a grid pattern. Therefore, the photographed image of the measurement range reflects not only the reinforcing bars 5 in the plane closest to the photographing unit 2, but also the reinforcing bars 5 in the plane behind it (lower layer).

[0052] Therefore, the reinforcement detection unit 15 determines, as the measurement target, the rebar 5 that straddles the tape 6 among the multiple rebars 5 that appear in the captured image of the measurement range, and calculates rebar position information for the determined rebar 5. For example, the reinforcement detection unit 15 performs image analysis on the captured image of the measurement range to identify the image of the rebar 5 that straddles the tape 6. Then, the reinforcement detection unit 15 calculates two-dimensional coordinate information that indicates the position of the identified rebar 5 in the captured image, and outputs the calculated two-dimensional coordinate information to the measurement processing unit 16.

[0053] The measurement processing unit 16 measures the reinforcing bars 5 using the reinforcing bar position information detected by the reinforcing bar arrangement detection unit 15 (step ST6). For example, the measurement processing unit 16 uses the reinforcing bar position information indicating the position of each of the multiple reinforcing bars 5 in the measurement range to calculate the number of reinforcing bars 5, the average reinforcing bar arrangement spacing of the reinforcing bars 5, the diameter of the reinforcing bars 5, and the spacing between nodes of the reinforcing bars 5 for each measurement range. The measurement processing unit 16 outputs display information indicating the identification number of the measurement range and the measurement results of the reinforcing bars 5 in the measurement range to the display processing unit 17. The display processing unit 17 displays the measurement results of the reinforcing bars 5 by the measurement processing unit 16 on the display unit 3.

[0054] For example, the display processing unit 17 generates display information for an electronic blackboard that lists the identification number for each measurement range and the measurement results of the rebars 5, and displays the generated display information superimposed on the image on the window screen M2 shown in Fig. 7. Furthermore, when the measurement processing unit 16 has completed rebar measurements in all measurement ranges in the inspection area 4, it tallies the number of rebars for each measurement range and calculates the average rebar spacing for each measurement range. The measurement results, including the tallied number of rebars, the average rebar spacing, and the diameters and spacing of the rebars 5 in all measurement ranges, are displayed on the display unit 3 by the display processing unit 17.

[0055] Furthermore, if the inspection area 4 is a wide area in the depth direction as seen from the imaging unit 2, that is, a wide area also in the direction perpendicular to the movement direction of the reinforcement bar inspection device 1 on a plane parallel to the horizontal plane, the process of specifying the measurement range is also performed in the depth direction. For example, after completing measurement of the rebars 5 in the initial movement direction, the inspector repositions the tape 6 that was laid in the previous movement direction in the depth direction of the inspection area 4. For example, the inspector repositions the tape 6 in the depth direction using the position of a specific marker 7 (for example, a marker 7 at the end of the inspection area 4) among the multiple markers 7 on the tape 6 as a reference. The inspector moves the reinforcement bar inspection device 1 along the repositioned tape 6, thereby performing measurements according to the procedure described above.

[0056] FIG. 8 is a flowchart showing the details of the measurement process of the reinforcing bar 5, and shows the details of the process of step ST6 in FIG. The image conversion unit 161 converts the captured image of the measurement range in the inspection area 4 acquired by the captured image acquisition unit 14 into an oriented image (step ST1A). FIG. 9 is an explanatory diagram showing the process of converting the captured image 2C into an oriented image 2D. The captured image 2C is a captured image of the measurement range. In the captured image 2C and the oriented image 2D in FIG. 9, only the rebars 5 in the top layer, i.e., the plane closest to the imaging unit 2, of the multiple planes that make up the inspection area 4 are depicted.

[0057] 9, in photographed image 2C, the closer the rebar 5 is to photographing unit 2, the larger it appears, and the further it is from photographing unit 2, the smaller it appears. That is, the lower image area in photographed image 2C shows rebar 5 that is closer to photographing unit 2, and the upper image area shows rebar 5 that is farther from photographing unit 2.

[0058] The image conversion unit 161 specifies the four corner points of any rectangle among the multiple rebars 5 shown in the captured image 2C, and estimates a homography transformation matrix that gives the rectangle with the specified four corner points a shape as seen from the front of the imaging unit 2. Next, the image conversion unit 161 converts the captured image 2C into a normal-oriented image 2D using the homography transformation matrix. Furthermore, the image conversion unit 161 corrects the reinforcing bar position information detected from the captured image 2C by the reinforcing bar arrangement detection unit 15 to match it with the orientated image 2D. As a result, the reinforcing bar position information is corrected to information indicating the position of the reinforcing bar 5 in the orientated image 2D.

[0059] All points (pixels) in the orthogonal image 2D are scaled so that the distance from the photographing unit 2 is constant. Therefore, in the orthogonal image 2D, the difference in size of the reinforcing bars 5 according to the distance between the photographing unit 2 and the plane of the measurement range is corrected.

[0060] The image extraction unit 162 extracts an image of the reinforcing bar 5 from the orientated image 2D based on the orientated image 2D and the reinforcing bar position information (step ST2A). For example, the image extraction unit 162 extracts an image of the reinforcing bar 5 in the orientated image 2D based on the reinforcing bar position information, and outputs the extracted image to the measurement unit 163.

[0061] The measurement unit 163 measures the rebars 5 using the images of the rebars 5 extracted by the image extraction unit 162 (step ST3A). For example, the measurement unit 163 measures the number of rebars 5 (number of arranged rebars) in the measurement range based on the number of images of the rebars 5 extracted from the orthogonal image 2D. The measurement unit 163 also measures the spacing between the rebars 5 on the orthogonal image 2D based on the rebar position information of adjacent rebars 5. The measurement unit 163 also performs image analysis on the images of the rebars 5 to measure the diameter and spacing of the nodes of the rebars 5. For example, the measurement unit 163 enhances the contours of the rebars 5 on the image by, for example, binarizing the image of the rebar 5, and measures the width of the enhanced contour as the diameter of the rebar 5. Similarly, the measurement unit 163 enhances the nodes of the rebars 5 on the image by, for example, binarizing the image of the rebar 5, and measures the spacing between the enhanced nodes as the node spacing of the rebars 5.

[0062] There are 16 types of rebars based on the JIS standard. These 16 types of rebars have "designations" such as D4, D5, D6, D8, D10, D13, D16, D19, D22, D25, D29, D32, D35, D38, D41, and D51. The designation indicates the rounded diameter of the rebar's nominal diameter. For example, the nominal diameter of a D10 rebar is 9.53 mm, the nominal diameter of a D13 rebar is 12.7 mm, and the nominal diameter of a D16 rebar is 15.9 mm. Rebars generally used as the framework of buildings are D10 or higher.

[0063] Furthermore, each of the 16 types of rebar has a specified "maximum average spacing between knots." For example, the maximum average spacing between knots for D10 rebars is 6.7 mm, the maximum average spacing between knots for D13 rebars is 8.9 mm, and the maximum average spacing between knots for D16 rebars is 11.1 mm.

[0064] As mentioned above, the diameters of the 16 types of rebar increase in increments of approximately 3 mm. Since a difference of 3 mm between the rebars in photographed image 2C generally amounts to only a difference of a few pixels, a high-resolution photographed image 2C is required to detect the actual diameter of the rebar from the outline width of the rebar in photographed image 2C. Furthermore, since only the maximum value of the average spacing between knots is specified for rebars, the spacing between knots in actual rebars varies depending on the manufacturer or production lot.

[0065] Therefore, the measurement unit 163 may use a learning model to infer at least one of the diameter and the node spacing of the reinforcing bar 5. By using the learning model, the measurement unit 163 can accurately detect the diameter or the node spacing of the reinforcing bar 5, and therefore can determine the type of the reinforcing bar 5 based on the diameter or the node spacing of the reinforcing bar 5.

[0066] 10 is a block diagram showing the configuration of the measurement unit 163 and the learning device 9. The learning device 9 receives as input a plurality of partial images extracted from the image of the reinforcing bar 5 in the 2D normalized image, and generates a learning model that outputs inspection information for the reinforcing bar 5.

[0067] The learning device 9 includes a learning unit 91 and a memory unit 92. The inspection information of the reinforcing bars 5 is at least one of the diameter of the reinforcing bars 5 and the spacing between nodes of the reinforcing bars 5. The learning unit 91 generates a learning model that has learned the inspection information of the reinforcing bars 5 using the learning data, and stores the learning model in the memory unit 92. As the learning algorithm, for example, deep learning, neural networks, genetic programming, functional logic programming, or support vector machines may be used.

[0068] The learning data is a dataset that includes multiple partial images and correct labels of reinforcing bars 5, which are input data for the learning model. The correct labels are information indicating the diameter of the reinforcing bars 5 corresponding to each of the multiple partial images, information indicating the node spacing of the reinforcing bars 5 corresponding to each of the multiple partial images, or information indicating the diameter and node spacing of the reinforcing bars 5 corresponding to each of the multiple partial images.

[0069] The image extraction unit 162 extracts multiple partial images from the image of the reinforcing bar 5 and outputs them to the measurement unit 163. The multiple partial images are obtained by dividing partial images of the reinforcing bar 5 of the same image size in sequence along the longitudinal direction of the reinforcing bar 5. The measurement unit 163 infers at least one of the diameter and the spacing between nodes of the reinforcing bar 5 using a learning model in the memory unit 92. The learning model is a machine learning model that receives as input the multiple partial images extracted from the 2D orientated image by the image extraction unit 162, and outputs at least one of the diameter and the spacing between nodes of the reinforcing bar 5 indicated by the partial images.

[0070] 11 is an explanatory diagram showing the process of extracting partial images 2F1, 2F2, 2F3, etc. of the rebar 5 from the orientated image 2D. The image extraction unit 162 extracts image 2E of the rebar 5 from the orientated image 2D. The image extraction unit 162 extracts partial images 2F1, 2F2, 2F3, etc. of the same image size from image 2E and outputs them to the measurement unit 163. The partial images 2F1, 2F2, 2F3, etc. are, for example, square images with the same number of pixels in both length and width.

[0071] The image extraction unit 162 extracts partial images of all the reinforcing bars 5 in the orthogonal image 2D. The measurement unit 163 inputs the partial images of the reinforcing bars 5 extracted from the orthogonal image 2D into a learning model, thereby inferring at least one of the diameter D of the reinforcing bars 5 and the spacing L of the nodes of the reinforcing bars 5.

[0072] FIG. 12 is an explanatory diagram showing the process of inputting a partial image of a rebar 5 into a learning model to infer the diameter D of the rebar. In FIG. 12, the learning model is a model generated using learning data, which is a set of multiple partial images of the rebar 5 and correct labels indicating the diameter D of the rebar assigned to each partial image. The measurement unit 163 inputs multiple pieces of partial image data into the learning model, and performs inference for each piece of partial image data. As a result, the learning model calculates, for each partial image, the probability that a rebar 5 with diameter D is captured and the probability that something other than a rebar is captured.

[0073] For example, inference result data 1631 is the result of calculating the probability that each type of rebar appears in partial image 2F1 shown in FIG. 11, and the probability that something other than rebar (NON) appears in the image. D10 to D51 in inference result data 1631 are the names of rebars. Similarly, inference result data 1632 is the result of calculating the probability that each type of rebar appears in partial image 2F2 shown in FIG. 11, and the probability that something other than rebar (NON) appears in the image. Such inference result data is calculated for all partial image data extracted from the same rebar image.

[0074] Next, the learning model calculates inference result data 164 for the image of the rebar by averaging the inference result data for all partial image data extracted from the same image of the rebar. For example, the inference result data 164 includes an average value of the probability that each type of rebar is captured and an average value of the probability that something other than a rebar (NON) is captured. The measurement unit 163 determines the diameter D of the rebar based on the inference result data 164. In the example of FIG. 12, the type of rebar 5 in image 2E shown in FIG. 11 is determined to be D16, i.e., a rebar 5 with a diameter D of 15.9 (mm). By using the above learning model, the measurement unit 163 can accurately detect the diameter D of the rebar and determine the type of rebar 5 based on the diameter D of the rebar.

[0075] In addition, the measurement unit 163 may infer the spacing L of the nodes of the reinforcing bars 5 using a learning model that takes multiple partial images as input and outputs the spacing L of the nodes of the reinforcing bars 5 in an image of the reinforcing bars 5 extracted from the orientated image 2D.

[0076] The learning model is a model generated using learning data that is a set of a plurality of partial images of the reinforcing bar 5 and correct labels indicating the spacing L of the nodes of the reinforcing bar 5 assigned to each partial image. The measurement unit 163 inputs multiple partial image data into a learning model and performs inference for each partial image data, calculating, for each partial image, the probability that a reinforcing bar 5 with a node spacing of L is captured and the probability that something other than a reinforcing bar is captured.

[0077] The learning model calculates the inference result data for the image of the rebar 5 by averaging the inference result data for all partial image data extracted from the same image of the rebar 5. Based on the calculated inference result data, the measurement unit 163 can accurately detect the spacing L of the nodes of the rebar 5 and can determine the type of the rebar 5 based on the spacing L of the nodes of the rebar 5.

[0078] Furthermore, the measurement unit 163 may infer the diameter D and the spacing L of the reinforcing bars 5 as inspection information of the reinforcing bars 5 using a learning model that takes multiple partial images as input and outputs the diameter D and the spacing L of the reinforcing bars in the image of the reinforcing bars extracted from the orthogonal image 2D.

[0079] The learning model is a model generated using learning data that is a set of multiple partial images of rebars 5 and correct labels assigned to each partial image that indicate the diameter D and node spacing L of the rebars 5. The measurement unit 163 inputs multiple partial image data into the learning model, performs inference for each partial image data, and calculates, for each partial image, the probability that a rebar 5 with diameter D and node spacing L is captured and the probability that something other than a rebar is captured.

[0080] The learning model calculates the inference result data for the image of the rebar 5 by averaging the inference result data for all partial image data extracted from the same image of the rebar 5. Based on the calculated inference result data, the measurement unit 163 can accurately detect the diameter D and the spacing L of the nodes of the rebar 5, and can determine the type of the rebar 5 based on the diameter D and the spacing L of the nodes of the rebar 5.

[0081] The functions of the identifier detection unit 11, identifier position information calculation unit 12, measurement range specification unit 13, photographed image acquisition unit 14, reinforcement detection unit 15, measurement processing unit 16, and display processing unit 17 provided in the reinforcement inspection device 1 are realized by a processing circuit. That is, the reinforcement inspection device 1 includes a processing circuit for executing the processes of steps ST1 to ST6 shown in Fig. 3. The processing circuit may be dedicated hardware, or may be a CPU (Central Processing Unit) that executes a program stored in memory.

[0082] Fig. 13A is a block diagram showing a hardware configuration that realizes the functions of the bar arrangement inspection apparatus 1. Fig. 13B is a block diagram showing a hardware configuration that executes software that realizes the functions of the bar arrangement inspection apparatus 1. In Figs. 13A and 13B, an input interface 100 is an interface that relays captured image data output from the imaging unit 2 to the bar arrangement inspection apparatus 1. An output interface 101 is an interface that relays display information output from the bar arrangement inspection apparatus 1 to the display unit 3.

[0083] When the processing circuit is the dedicated hardware processing circuit 102 shown in FIG. 13A, the processing circuit 102 may be, for example, a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC (Application Specific Integrated Circuit), an FPGA (Field-Programmable Gate Array), or a combination thereof. The functions of the identifier detection unit 11, identifier position information calculation unit 12, measurement range determination unit 13, captured image acquisition unit 14, reinforcement detection unit 15, measurement processing unit 16 and display processing unit 17 provided in the reinforcement inspection device 1 may be realized by separate processing circuits, or these functions may be realized together by a single processing circuit.

[0084] 13B, the functions of the identifier detection unit 11, identifier position information calculation unit 12, measurement range specification unit 13, photographed image acquisition unit 14, reinforcement detection unit 15, measurement processing unit 16, and display processing unit 17 included in the reinforcement inspection device 1 are realized by software, firmware, or a combination of software and firmware. The software or firmware is written as a program and stored in the memory 104.

[0085] The processor 103 reads and executes programs stored in the memory 104 to realize the functions of the identifier detection unit 11, the identifier position information calculation unit 12, the measurement range specification unit 13, the photographed image acquisition unit 14, the reinforcement arrangement detection unit 15, the measurement processing unit 16, and the display processing unit 17 included in the reinforcement arrangement inspection device 1. For example, the reinforcement arrangement inspection device 1 includes a memory 104 for storing programs that, when executed by the processor 103, result in the execution of steps ST1 to ST6 shown in FIG. 3. These programs cause a computer to execute the procedures or methods of the processes performed by the identifier detection unit 11, the identifier position information calculation unit 12, the measurement range specification unit 13, the photographed image acquisition unit 14, the reinforcement arrangement detection unit 15, the measurement processing unit 16, and the display processing unit 17. The memory 104 may be a computer-readable storage medium that stores programs that cause a computer to function as the identifier detection unit 11, the identifier position information calculation unit 12, the measurement range specification unit 13, the photographed image acquisition unit 14, the reinforcement arrangement detection unit 15, the measurement processing unit 16, and the display processing unit 17.

[0086] Memory 104 may be, for example, a non-volatile or volatile semiconductor memory such as RAM (Random Access Memory), ROM (Read Only Memory), flash memory, EPROM (Erasable Programmable Read Only Memory), EEPROM (Electrically-EPROM), a magnetic disk, a flexible disk, an optical disk, a compact disk, a mini disk, a DVD, etc.

[0087] The functions of the identifier detection unit 11, the identifier position information calculation unit 12, the measurement range specification unit 13, the photographed image acquisition unit 14, the reinforcement detection unit 15, the measurement processing unit 16, and the display processing unit 17 included in the reinforcement inspection device 1 may be partly realized by dedicated hardware, and the remaining part may be partly realized by software or firmware. For example, the identifier detection unit 11, the identifier position information calculation unit 12, and the measurement range specification unit 13 are

[0088] The functions are realized by a processing circuit 102, which is dedicated hardware, and the functions of the photographed image acquisition unit 14, the reinforcement detection unit 15, the measurement processing unit 16, and the display processing unit 17 are realized by a processor 103 reading and executing programs stored in a memory 104. In this way, the processing circuit can realize the above functions by hardware, software, firmware, or a combination of these.

[0089] 10 shows the learning device 9 provided separately from the reinforcement bar arrangement inspection device 1, the learning unit 91 and the memory unit 92 may be provided in the reinforcement bar arrangement inspection device 1. The memory unit 92 may also be provided in an external device capable of data communication with the reinforcement bar arrangement inspection device 1.

[0090] The memory unit 92 sequentially stores partial image data extracted by the image extraction unit 162 during reinforcement inspection. The partial image data stored in the memory unit 92 are assigned correct answer labels input using, for example, an operation unit (not shown in FIG. 1 ) and stored in the memory unit 92 as learning data. The learning unit 91 generates a learning model using the learning data stored in the memory unit 92. The measurement unit 163 infers inspection information for the reinforcing bars 5 using the learning model generated by the learning unit 91. The measurement unit 163 may also use a learning model generated for each site. In this case, the accuracy of reinforcement inspection for each site is improved.

[0091] As described above, the reinforcement bar inspection device 1 according to embodiment 1 includes an identifier detection unit 11 that acquires an image of the inspection area 4 and the tape 6 from the photographing unit 2 and detects the marker 7 from the acquired image, an identifier position information calculation unit 12 that calculates identifier position information indicating the position of the marker 7, a measurement range determination unit 13 that determines the measurement range of the reinforcing bar 5 using the identifier position information, an image acquisition unit 14 that acquires photographed image data of the measurement range from the photographing unit 2, a reinforcement bar detection unit 15 that uses the photographed image data of the measurement range to detect reinforcing bar position information indicating the position of the reinforcing bar 5 spanning the tape 6, and a measurement processing unit 16 that measures the reinforcing bar 5 corresponding to the reinforcing bar position information. The reinforcing bar inspection device 1 automatically specifies the measurement range of the reinforcing bars 5 using the identifier position information, so the measurement range of the reinforcing bars 5 can be specified from the inspection area 4 without relying on visual inspection by an inspector. In addition, the reinforcing bar inspection device 1 determines that the reinforcing bars 5 that are straddling the tape 6 in the captured image 2C are the measurement target, so it can identify the reinforcing bars 5 that are the measurement target without analyzing the three-dimensional structure of the inspection area 4.

[0092] In the bar arrangement inspection device 1 according to the first embodiment, a plurality of markers 7 are provided at equal intervals on the tape 6. Based on the positions of the markers 7 provided at equal intervals in the longitudinal direction of the tape 6, the bar arrangement inspection device 1 can identify positions where the measurement range of the inspection area 4 can be photographed.

[0093] In the bar arrangement inspection device 1 according to the first embodiment, the captured image acquisition unit 14 allows the imaging unit 2 to capture an image of the measurement range when the measurement range is identified from the image of the inspection area 4 where the tape 6 is placed. This prevents the inspector from mistakenly capturing an image in which the measurement range is not identified.

[0094] In the reinforcement bar arrangement inspection device 1 according to the first embodiment, the captured image acquisition unit 14 automatically causes the photographing unit 2 to photograph the measurement range when the measurement range is identified from the video of the inspection area 4 where the tape 6 is placed. This allows the inspector to obtain photographed image data for each measurement range simply by moving the reinforcement bar arrangement inspection device 1 along the tape 6.

[0095] The reinforcement bar arrangement inspection device 1 according to the first embodiment includes a display processing unit 17 that displays an image of the inspection area 4 where the tape 6 is placed on the display unit 3 in different modes until the measurement range is identified and when the measurement range has been identified. By visually checking the display mode of the image of the inspection area 4, the inspector can easily understand whether the measurement range has been identified or not.

[0096] In the bar arrangement inspection device 1 according to the first embodiment, the measurement processing unit 16 includes an image conversion unit 161, an image extraction unit 162, and a measurement unit 163. The image conversion unit 161 converts the captured image 2C into an oriented image 2D. The image extraction unit 162 extracts an image of the reinforcing bar 5 from the oriented image 2D. The measurement unit 163 measures the reinforcing bar 5 using the image of the reinforcing bar 5. This enables the bar arrangement inspection device 1 to accurately measure the reinforcing bar 5 within the measurement range.

[0097] In the bar arrangement inspection device 1 according to the first embodiment, the measurement unit 163 receives as input an image of the reinforcing bar 5 extracted from the orientated image 2D, and infers inspection information of the reinforcing bar 5 within the measurement range using a learning model that outputs inspection information of the reinforcing bar 5. By using the learning model, the measurement unit 163 can accurately measure the inspection information of the reinforcing bar 5.

[0098] In the reinforcement bar arrangement inspection device 1 according to the first embodiment, the image extraction unit 162 extracts a plurality of partial images from the image of the reinforcing bar 5 extracted from the orientated image 2D. The measurement unit 163 inputs the plurality of partial images and infers the diameter D of the reinforcing bar 5 using a learning model that outputs the diameter D of the reinforcing bar 5 indicated by the partial images as inspection information of the reinforcing bar 5. By using the learning model, the measurement unit 163 can accurately measure the diameter D of the reinforcing bar 5.

[0099] In the reinforcement bar arrangement inspection device 1 according to the first embodiment, the image extraction unit 162 extracts a plurality of partial images from the image of the reinforcing bar 5 extracted from the orientated image 2D. The measurement unit 163 receives the plurality of partial images and infers the spacing L of the nodes of the reinforcing bar 5 using a learning model that outputs the spacing L of the nodes of the reinforcing bar 5 indicated by the partial images as inspection information of the reinforcing bar 5. By using the learning model, the measurement unit 163 can accurately measure the spacing L of the nodes of the reinforcing bar 5.

[0100] In the reinforcement bar arrangement inspection device 1 according to the first embodiment, the image extraction unit 162 extracts a plurality of partial images from the image of the reinforcing bar 5 extracted from the orientated image 2D. The measurement unit 163 inputs the plurality of partial images and infers the diameter D and the spacing L of the reinforcing bar 5 using a learning model that outputs the diameter D and the spacing L of the nodes of the reinforcing bar 5 indicated by the partial images as inspection information of the reinforcing bar 5. By using the learning model, the measurement unit 163 can accurately measure the diameter D and the spacing L of the nodes of the reinforcing bar 5.

[0101] Any of the components of the embodiments may be modified or omitted. [Explanation of symbols]

[0102] 1 Reinforcement inspection device, 2 Photographing unit, 2A, 2E images, 2B Photographing operation image, 2C Photographed image, 2D oriented image, 2F1 to 2F3 partial images, 3 Display unit, 3A, 3B display screen, 4 Inspection area, 5 Reinforcement bar, 6 Tape, 7 Marker, 7A Base, 7B Identification shape, 8 Measurement range, 9 Learning device, 11 Identifier detection unit, 12 Identifier position information calculation unit, 13 Measurement range identification unit, 14 Photographed image acquisition unit, 15 Reinforcement detection unit, 16 Measurement processing unit, 17 Display processing unit, 91 Learning unit, 92 Memory unit, 100 Input interface, 101 Output interface, 102 Processing circuit, 103 Processor, 104 Memory, 161 Image conversion unit, 162 Image extraction unit, 163 Measurement unit, 164 Inference result data, 1631, 1632 Inference result data.

Claims

1. an identifier detection unit that acquires from a photographing unit images of an inspection area in which a plurality of reinforcing bars are arranged, and a strip-shaped member on which a plurality of identifiers are provided at intervals in the longitudinal direction and which is placed under the reinforcing bars to be measured in the inspection area, as time-series data of frame images taken by a camera moving along the strip-shaped member, and detects the identifiers from each of the frame images of the acquired images; an identifier position information calculation unit that calculates identifier position information indicating a position on the frame image of the identifier detected by the identifier detection unit when it is determined that the frame image in which the identifier is detected corresponds to an image obtained by photographing the inspection area from the front; and a measurement range specifying unit that specifies a measurement range of the reinforcing bar using the identifier position information, and when two adjacent identifiers are located in one of the frame images, specifies a measurement range of the reinforcing bar using the identifier position information for the two identifiers, and outputs position information of the specified measurement range; a captured image acquisition unit that acquires position information of the measurement range output by the measurement range identification unit when the measurement range is identified, outputs a shooting instruction signal to the imaging unit, and acquires captured image data of the measurement range captured based on the shooting instruction signal from the imaging unit; a reinforcing bar arrangement detection unit that detects reinforcing bar position information indicating the position of the reinforcing bar spanning over the strip-shaped member using photographed image data of the measurement range; a measurement processing unit that measures the reinforcing bar corresponding to the reinforcing bar position information. A reinforcement inspection device characterized by the above.

2. The band-shaped member has a plurality of the identifiers provided at equal intervals.

2. The reinforcing bar inspection device according to claim 1.

3. When the measurement range is identified from the image of the inspection area in which the strip-shaped member is arranged, the captured image acquisition unit permits the imaging unit to capture the measurement range by the instruction signal.

3. The reinforcing bar inspection device according to claim 1 or 2.

4. When the measurement range is identified from an image of the inspection area in which the strip-shaped member is arranged, the photographed image acquisition unit automatically causes the photographing unit to photograph the measurement range in response to the instruction signal.

3. The reinforcing bar inspection device according to claim 1 or 2.

5. a display processing unit that displays an image of the inspection area in which the strip-shaped member is arranged on a display unit in a manner different from that before the measurement range is specified and when the measurement range is specified; 5. The reinforcing bar inspection device according to claim 3 or 4.

6. The measurement processing unit an image conversion unit that converts the captured image of the measurement range into a normal-oriented image in which the distance between the image capture unit and the measurement range is constant and the measurement range is normal to the image capture unit; an image extraction unit that extracts an image of the reinforcing bar corresponding to the reinforcing bar position information from the orthogonal image; a measuring unit that measures the reinforcing bar using the image of the reinforcing bar extracted from the orthogonal image. The bar arrangement inspection device according to any one of claims 1 to 5.

7. The measurement unit receives an image of the reinforcing bar extracted from the orientated image as an input, and infers inspection information of the reinforcing bar within the measurement range using a learning model that outputs inspection information of the reinforcing bar.

7. The bar arrangement inspection device according to claim 6.

8. The image extraction unit extracts a plurality of partial images from the image of the reinforcing bar extracted from the orthogonal image, The measurement unit inputs the plurality of partial images and infers the diameter of the reinforcing bar using the learning model that outputs the diameter of the reinforcing bar indicated by the partial images as inspection information of the reinforcing bar. The bar arrangement inspection device according to claim 7 .

9. The measurement unit inputs the plurality of partial images and infers the spacing of the nodes of the reinforcing bar using the learning model that outputs the spacing of the nodes of the reinforcing bar indicated by the partial images as inspection information of the reinforcing bar. The bar arrangement inspection device according to claim 8 .

10. The photographing unit is a monocular camera. The bar arrangement inspection device according to any one of claims 1 to 9.

11. The identifier detection unit acquires from the photographing unit images of an inspection area in which a plurality of reinforcing bars are arranged, and a strip-shaped member on which a plurality of identifiers are provided at intervals in the longitudinal direction and which is placed under the reinforcing bars to be measured in the inspection area, the image being captured as time-series data of frame images by a camera moving along the strip-shaped member, and detects the identifiers from each of the frame images of the acquired images; a step of calculating identifier position information indicating a position on the frame image of the identifier detected by the identifier detection unit when the identifier position information calculation unit determines that the frame image in which the identifier is detected corresponds to an image obtained by photographing the inspection area from the front; a step in which a measurement range specifying unit specifies, when two adjacent identifiers are located in one frame image, a measurement range of the reinforcing bar by using the identifier position information for the two identifiers and outputting position information of the specified measurement range; a step in which a photographed image acquisition unit acquires position information of the measurement range output by the measurement range identification unit when the measurement range is identified, outputs a photographing instruction signal to the photographing unit, and acquires photographed image data of the measurement range photographed based on the photographing instruction signal from the photographing unit; a step in which a reinforcement arrangement detection unit detects reinforcing bar position information indicating the position of the reinforcing bar spanning over the belt-shaped member using the photographed image data of the measurement range; a step in which a measurement processing unit measures the reinforcing bar corresponding to the reinforcing bar position information; A reinforcement inspection method characterized by the above.

12. Computer, an identifier detection unit that acquires from a photographing unit images of an inspection area in which a plurality of reinforcing bars are arranged, and a strip-shaped member on which a plurality of identifiers are provided at intervals in the longitudinal direction and which is placed under the reinforcing bars to be measured in the inspection area, the strip-shaped member being photographed as time-series data of frame images by a camera moving along the strip-shaped member, and detects the identifiers from each of the frame images of the acquired images; an identifier position information calculation unit that calculates identifier position information indicating the position of the identifier detected by the identifier detection unit on the frame image when it is determined that the frame image in which the identifier is detected corresponds to an image obtained by photographing the inspection area from the front; a measurement range specifying unit that specifies a measurement range of the reinforcing bar using the identifier position information, and when two adjacent identifiers are located in one of the frame images, specifies a measurement range of the reinforcing bar using the identifier position information for the two identifiers and outputs position information of the specified measurement range; a captured image acquisition unit that acquires position information of the measurement range output by the measurement range identification unit when the measurement range is identified, outputs a shooting instruction signal to the imaging unit, and acquires, from the imaging unit, captured image data of the measurement range captured based on the shooting instruction signal; a reinforcing bar arrangement detection unit that detects reinforcing bar position information indicating the positions of the reinforcing bars straddling the belt-shaped member using the photographed image data of the measurement range; a measurement processing unit that measures the reinforcing bar corresponding to the reinforcing bar position information; A program to function as a

Images