Bar arrangement inspection device, bar arrangement inspection method, and program

The reinforcement inspection device addresses the challenge of measuring reinforcing bars on curved surfaces by using stereo imaging and partial plane conversion to accurately detect and measure reinforcing bar dimensions on lattice-shaped structures.

JP2024167464APending Publication Date: 2024-12-04MITSUBISHI ELECTRIC ENG CO LTD
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
JP2023083562
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Filing Date
2023-05-22
Publication Date
2024-12-04

AI Technical Summary

Technical Problem

Conventional reinforcement inspection devices are unable to accurately measure reinforcing bars on curved surfaces due to the distortion and occlusion caused by the curvature, which affects the detection and measurement of reinforcing bars arranged in a lattice shape.

Method used

The reinforcement inspection device uses a stereo camera to generate three-dimensional image data, sets partial planes along the curved surface, converts these planes into orthogonal front-facing images, and associates reinforcing bars across multiple images to accurately measure their positions and dimensions.

Benefits of technology

Enables precise measurement of reinforcing bars on curved surfaces by correcting for distortion and occlusion, allowing for accurate detection and measurement of reinforcing bar diameters and spacing on lattice-shaped arrangements.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2024167464000001_ABST
    Figure 2024167464000001_ABST
Patent Text Reader

Abstract

To provide a bar arrangement inspection device that can measure reinforcing bars of a bent surface comprising the reinforcing bars assembled in a lattice shape.SOLUTION: A bar arrangement inspection device (1) comprises: a three-dimensional image data acquisition unit (11) that generates three-dimensional image data of a curved surface by using a stereoscopic image of bar arrangement; a partial plane setting unit (12) that sets a plurality of partial planes along the curved surface; an image conversion unit (13) that converts images of the partial planes into facing images in which reinforcing bars (6) are orthogonal to each other; a reinforcing bar detection unit (14) that detects the positions of the reinforcing bars (6) by using the facing images; an association unit (15) that associates the same reinforcing bars (6) with each other, which are detected in the plurality of facing images; and a measurement processing unit (16) that measures the reinforcing bars (6) on the curved surface on the basis of information on the positions of the reinforcing bars (6) associated as being the same.SELECTED DRAWING: Figure 1
Need to check novelty before this filing date? Find Prior Art

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, a reinforcing bar inspection is performed to check whether the reinforcing bars, which are assembled together, are constructed as designed. For example, Patent Document 1 describes a reinforcing bar inspection device that inspects reinforcing bars having a multi-layered plane made of reinforcing bars assembled in a lattice pattern. This reinforcing bar inspection device converts three-dimensional image data of reinforcing bars photographed by a stereo camera into a frontal image, detects the position of the reinforcing bars on the plane to be inspected from the frontal image, extracts an image of the reinforcing bars from the frontal image based on the position of the reinforcing bars, and inspects the reinforcing bars using the image of the reinforcing bars. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2022-164949 A Summary of the Invention [Problem to be solved by the invention]

[0004] The conventional reinforcing bar inspection device described in Patent Document 1 extracts the plane of reinforcing bars arranged in a lattice pattern from the reinforcing bars and measures the reinforcing bars in that plane, so there was a problem in that it was not possible to measure reinforcing bars with curved surfaces consisting of reinforcing bars arranged in a lattice pattern.

[0005] The present disclosure is devised to solve the above-mentioned problems, and aims to provide a reinforcing bar inspection device capable of measuring reinforcing bars on a curved surface that are arranged in a lattice pattern. [Means for solving the problem]

[0006] The reinforcement inspection device of the present disclosure is a reinforcement inspection device that inspects reinforcement having a curved surface consisting of rebars arranged in a lattice pattern, and includes a three-dimensional image data acquisition unit that generates three-dimensional image data of the curved surface using stereo images of the reinforcement, a partial plane setting unit that sets multiple partial planes along the curved surface in the stereo images, an image conversion unit that converts the images of the partial planes into front-facing images in which the rebars are perpendicular to each other, a rebar detection unit that detects the positions of the rebars included in the partial planes using the front-facing images, a matching unit that matches identical rebars detected in the multiple front-facing images, and a measurement processing unit that measures the rebars on the curved surface based on rebar position information of the rebars that have been matched as identical. Effect of the Invention

[0007] According to the present disclosure, multiple partial planes are set along the curved surfaces of stereo images of a curved surface made of reinforcing bars arranged in a lattice pattern, the images of the partial planes are converted into front-view images in which the reinforcing bars are perpendicular to each other, and the reinforcing bars included in the partial planes are detected using the front-view images. Then, identical reinforcing bars detected in the multiple front-view images are matched, and the reinforcing bars on the curved surface are measured based on the reinforcing bar position information of the reinforcing bars that are matched as identical. In this way, the reinforcing bar inspection device according to the present disclosure can measure the reinforcing bars on a curved surface made of reinforcing bars arranged in a lattice pattern. [Brief description of the drawings]

[0008] [Figure 1] 1 is a block diagram showing a configuration example of a bar arrangement inspection device according to a first embodiment. FIG. [Diagram 2] 2A, 2B, and 2C are schematic diagrams showing an overview of the reinforcing bar detection process. [Diagram 3] 1 is a screen diagram showing a captured image of a concave curved surface and a partial planar image relative to the stereo camera; [Figure 4] FIG. 1 is a schematic diagram showing an outline of a partial plane set on a curved surface. [Diagram 5] 11 is a screen diagram showing a captured image and a partial planar image of a curved surface of a convex reinforcement bar viewed from a stereo camera. FIG. [Figure 6]FIG. 13 is a schematic diagram showing an overview of a process for determining a reference position of a reinforcing bar. [Figure 7] 4 is a flowchart showing a bar arrangement inspection method according to the first embodiment. [Figure 8] 8A and 8B 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 PREFERRED EMBODIMENTS

[0009] Embodiment 1 FIG. 1 is a block diagram showing a configuration example of a reinforcing bar inspection device 1 according to a first embodiment. In FIG. 1, the reinforcing bar inspection device 1 is a device that inspects at least one of the number of reinforcing bars, the spacing between adjacent reinforcing bars, the diameter of the reinforcing bars, and the spacing between nodes of the reinforcing bars in a reinforcing bar arrangement having a curved surface made of reinforcing bars arranged in a lattice pattern. The reinforcing bar inspection device 1 is a tablet terminal, a smartphone, or a notebook-type personal computer (PC). As shown in FIG. 1, a stereo camera 2 is connected to the reinforcing bar inspection device 1 by wire or wirelessly. The reinforcing bar inspection device 1 is not limited to a device with an external stereo camera 2, and may have the stereo camera 2 built-in. In a reinforcing bar inspection, an inspector photographs the reinforcing bars with the stereo camera 2.

[0010] The bar arrangement inspection device 1 may be a component included in a server capable of communicating with a terminal device equipped with a stereo camera 2. For example, the terminal device is capable of performing a bar arrangement inspection provided in the form of SaaS (Software as a Service). When performing a bar arrangement inspection in the form of SaaS, a bar arrangement inspection application does not need to be installed in the terminal device. The bar arrangement inspection application is executed on the server, and the terminal device is provided with measurement result information on a general-purpose web browser. The bar arrangement inspection application is stored in a storage unit included in the server. Furthermore, a reinforcing bar inspection application may be installed in the terminal device. In the terminal device in which the reinforcing bar inspection application is installed, the application is executed, thereby enabling the reinforcing bar inspection.

[0011] The stereo camera 2 is an imaging device having a left imaging section and a right imaging section. The left imaging section captures a left viewpoint image seen from the left side. The right imaging section captures a right viewpoint image seen from the right side. The stereo camera 2 outputs a captured image composed of the left viewpoint image and the right viewpoint image. The camera angle of view information (H, V) for each pixel in the shooting range of the stereo camera 2 is included in the stereo image and output to the bar arrangement inspection device 1. This camera angle of view information (H, V) is information indicating where each pixel constituting the stereo image is located in the camera's angle of view, and is information expressed as angles in the horizontal direction (H) and vertical direction (V) corresponding to the xy coordinates of the pixel.

[0012] The bar arrangement inspection device 1 includes a memory unit 3, an operation unit 4, and a display unit 5, and further includes a calculation unit not shown in Fig. 1. The memory unit 3 is a storage device included in a computer that functions as the bar arrangement inspection device 1, and includes storage such as a hard disk drive (HDD) or a solid state drive (SSD), or a memory 104 in Fig. 8B described later. The memory unit 3 may be provided outside the bar arrangement inspection device 1 as long as it is accessible by the bar arrangement inspection device 1.

[0013] The operation unit 4 is an input device that accepts operations for the screen display of the display unit 5 provided in the computer functioning as the bar arrangement inspection apparatus 1. For example, if the bar arrangement inspection apparatus 1 is a smartphone or a tablet terminal, the operation unit 4 is a touch panel provided integrally with the display screen of the display unit 5. If the bar arrangement inspection apparatus 1 is a PC, the operation unit 4 is, for example, a mouse or a keyboard.

[0014] The display unit 5 is a display device included in a computer that functions as the bar arrangement inspection device 1. For example, the display unit 5 is an LCD (Liquid Crystal Display) or an organic EL (Electroluminescence) display device.

[0015] Furthermore, the calculation unit controls the overall operation of the reinforcing bar inspection device 1. The calculation unit includes a three-dimensional image data acquisition unit 11, a partial plane setting unit 12, an image conversion unit 13, a reinforcing bar detection unit 14, a matching unit 15, and a measurement processing unit 16. The calculation unit executes the reinforcing bar inspection application to realize various functions of the three-dimensional image data acquisition unit 11, the partial plane setting unit 12, the image conversion unit 13, the reinforcing bar detection unit 14, the matching unit 15, and the measurement processing unit 16.

[0016] Here, a process for detecting reinforcing bars from a plane consisting of reinforcing bars arranged in a lattice pattern will be described, and then a process for detecting reinforcing bars from a curved surface consisting of reinforcing bars arranged in a lattice pattern will be described. 2A is a schematic diagram showing an overview of the process of detecting reinforcing bars from a plane 21 on which the reinforcing bars are arranged in a lattice pattern, and shows the stereo camera 2 and the reinforcing bars that are the subject of its image as viewed from the side. In Fig. 2A, the stereo camera 2 is placed in a position directly facing the plane 21, and captures an image of the plane 21 from this position.

[0017] By using images (stereo images) of the plane 21 captured by the stereo camera 2 in the field of view 2A, rebars on the plane 21 within the detection range A are detected. The detection range A is a three-dimensional spatial region with a distance B between the stereo camera 2 side (front side) and the opposite side (rear side) via the plane 21. The conventional reinforcing bar inspection device described in Patent Document 1 detects the positions of the rebars present in the detection range A using the stereo images of the plane 21, and measures the rebar diameter, etc., of the rebars whose positions have been detected.

[0018] For example, a conventional reinforcing bar inspection device uses stereo images of a plane 21 to generate three-dimensional point cloud data consisting of a plurality of three-dimensional point data including distance information from three-dimensional points of an object including reinforcing bars in the detection range A to the stereo camera 2. The conventional reinforcing bar inspection device identifies the plane 21 included in the detection range A by applying the RANSAC (RANdom SAmple Consensus) method to the three-dimensional point cloud data acquired from the detection range A.

[0019] Furthermore, when the stereo camera 2 faces the plane 21 as shown in FIG. 2A, the stereo image of the plane 21 captured by the stereo camera 2 is a front-facing image in which there is no unevenness in the diameter of the rebars due to the perspective of the shooting position, and the vertical and horizontal rebars are captured perpendicular to each other. A conventional reinforcing bar inspection device can measure the diameter of the reinforcing bars and the spacing between adjacent reinforcing bars on plane 21 by measuring the reinforcing bars in this stereo image on a pixel-by-pixel basis.

[0020] Fig. 2B is a schematic diagram showing an overview of the process of detecting reinforcing bars from curved surface 22A where reinforcing bars are arranged in a lattice pattern, showing a side view of stereo camera 2 and the reinforcing bars that are the subject of its imaging. In Fig. 2B, curved surface 22A is in the same position as plane 21 shown in Fig. 2A, and stereo camera 2 is placed in the same position as in Fig. 2A, and it is assumed that curved surface 22A is imaged at this position. A conventional reinforcing bar inspection device detects reinforcing bars on curved surface 22A that are included in detection range A by using stereo images of curved surface 22A captured by stereo camera 2 in the field of view 2A.

[0021] However, on curved surface 22A, even if included in the field of view 2A, some rebars C fall outside of detection range A, and conventional bar arrangement inspection devices are unable to accurately detect rebars C. For example, as shown in FIG. 2B, in a stereo image taken of curved surface 22A that is concave with respect to stereo camera 2, rebars C that fall outside detection range A appear larger than the rebars within detection range A, and the image is not viewed head-on like in the case of flat surface 21. Furthermore, rebars C that appear large hide adjacent rebars from the stereo camera 2 side, making it impossible to accurately detect rebars from the stereo image.

[0022] Fig. 2C is a schematic diagram showing an overview of the process of detecting reinforcing bars from curved surface 22A on which multiple partial planes are set, and shows the stereo camera 2 and the reinforcing bars that are the subject of its imaging, as viewed from the side. In Fig. 2C, as in Fig. 2B, it is assumed that curved surface 22A is located in the same position as plane 21 shown in Fig. 2A, and that stereo camera 2 images curved surface 22A at the same position as in Fig. 2A.

[0023] As shown in Fig. 2B, the reinforcing bars on the curved surface 22A cannot be accurately detected within the detection range A, so the reinforcing bar inspection device 1 sets multiple partial planes (three partial planes in Fig. 2C) along the curved surface 22A and detects the reinforcing bars within detection ranges D1, D2, and D3 corresponding to the partial planes. The detection ranges D1, D2, and D3 are three-dimensional spatial regions with an interval E between the stereo camera 2 side (front side) and the opposite side (rear side) via each partial plane.

[0024] For example, the reinforcing bar inspection device 1 uses stereo images for each partial plane to generate three-dimensional point cloud data in the detection range D1, which is made up of a plurality of three-dimensional point data including distance information from three-dimensional points of an object including reinforcing bars in the detection range D1 to the stereo camera 2. Similarly, three-dimensional point cloud data in the detection range D2 and three-dimensional point cloud data in the detection range D3 are generated. The reinforcing bar inspection device 1 identifies partial planes included in the detection range D1 by applying the RANSAC method to the three-dimensional point cloud data in the detection range D1. Similarly, it identifies partial planes included in the detection range D2 and partial planes included in the detection range D3.

[0025] The reinforcing bar inspection device 1 converts stereo images of partial planes included in the detection ranges D1, D2, and D3 into frontal images in which the reinforcing bar diameters are uniform due to the perspective of the shooting positions and the vertical and horizontal reinforcing bars are perpendicular to each other. The reinforcing bar inspection device 1 can then measure the diameters of the reinforcing bars and the spacing between adjacent reinforcing bars in each partial plane by measuring the reinforcing bars in these frontal images in pixel units. This allows the reinforcing bar inspection device 1 to measure the reinforcing bars on curved surfaces that are made up of reinforcing bars arranged in a lattice pattern.

[0026] Next, the components of the bar arrangement inspection device 1 will be described. The three-dimensional image data acquisition unit 11 generates three-dimensional image data of the curved surface using the stereo images of the reinforcement. For example, the three-dimensional image data acquisition unit 11 generates three-dimensional image data of the reinforcement including the curved surface from the stereo images of the reinforcement captured by the stereo camera 2.

[0027] Moreover, as an inspector carrying the reinforcement inspection device 1 moves around the reinforcement to be inspected, the stereo camera 2 continuously captures images of the reinforcement from multiple shooting angles at a predetermined frame rate (e.g., 30 fps). The three-dimensional image data acquisition unit 11 sequentially inputs the captured images, which are time-series data of a left viewpoint image of the reinforcement and a right viewpoint image (frame image) of the same reinforcement, from the stereo camera 2. The three-dimensional image data acquisition unit 11 then performs stereo matching between the left viewpoint image and the right viewpoint image to acquire three-dimensional information, which is a three-dimensional image of the inspection area, for each captured image (for each frame image).

[0028] The three-dimensional image data acquisition unit 11 reconstructs three-dimensional information of the reinforcement using the three-dimensional information acquired for each captured image. For example, the three-dimensional images acquired for each captured image are captured from different angles by moving the stereo camera 2. The three-dimensional image data acquisition unit 11 inputs three-dimensional image data corresponding to each of a plurality of photographed images acquired at each predetermined examination cycle.

[0029] The three-dimensional image data acquisition unit 11 then extracts three-dimensional points of characteristic points (edges, etc.) of the reinforcing bars that make up the reinforcement from each piece of three-dimensional image data acquired during the inspection cycle, and specifies where the same three-dimensional points are located in each piece of three-dimensional image data. In this way, the three-dimensional image data acquisition unit 11 restores three-dimensional image data in which the three-dimensional point group of the reinforcement is specified for each inspection period.

[0030] The partial plane setting unit 12 sets a plurality of partial planes along a curved surface in a stereo image captured by the stereo camera 2. For example, the storage unit 3 stores the curvature R of the curved surface of the reinforcement to be inspected and the number of partial planes that divide the curved surface. The partial plane setting unit 12 identifies the curved surface using three-dimensional point cloud data included in the three-dimensional image data, and sets a circle with a radius equal to the inverse of the curvature R read from the storage unit 3 along the curved surface when viewed from the side. The partial plane setting unit 12 obtains two points where the circle and the curved surface intersect when viewed from the side, sets a partial plane passing through the two points, and converts the coordinates (for example, the coordinates of four vertices) that define the partial plane into two-dimensional coordinates on the stereo image. The partial plane setting unit 12 repeats this series of processes as many times as the number of partial planes read from the storage unit 3, thereby enabling the setting of a plurality of partial planes along a curved surface in a stereo image.

[0031] The number of partial planes that divide the curved surface may be a preset value as described above, but may also be calculated using the curvature R of the curved surface of the reinforcement being inspected. For example, the partial plane setting unit 12 may identify the upper and lower edges of a curved surface in a stereo image, calculate the distance between the upper and lower edges, and determine the number of partial planes that further divide this interval into equal intervals, or may determine the number of partial planes based on the density of three-dimensional points included in the calculated interval.

[0032] Moreover, the partial plane may be appropriately set by the user using the operation unit 4. For example, the partial plane setting unit 12 displays a stereo image of a reinforcement having a curved surface on the display unit 5, and sets a plurality of bounding boxes on the curved surface on the screen using the operation unit 4, thereby setting a partial plane within these bounding boxes. Then, the partial plane setting unit 12 identifies the coordinates (e.g., coordinates of four vertices) that define the partial plane set on the screen, and the three-dimensional point information corresponding to these coordinates. This makes it possible to determine the coordinates of the partial plane on the stereo image, and the corresponding three-dimensional point data (e.g., distance information).

[0033] The image conversion unit 13 converts the image of the partial plane into a front-facing image in which the rebars are perpendicular to each other. For example, the image conversion unit 13 extracts an image of the partial plane from a stereo image in which the rebar arrangement is photographed, based on the coordinates of four vertices that define the partial plane. Next, the image conversion unit 13 estimates a front-facing transformation matrix using the position coordinates of the four corners of the rectangle of the partial plane in which the rebars are combined, and converts the image of the partial plane into a front-facing image based on the front-facing transformation matrix. Since the front-facing transformation matrix differs for each partial plane captured in the stereo image, the image conversion unit 13 estimates a front-facing transformation matrix for each partial plane, and converts the image of the partial plane into a front-facing image using the estimated front-facing transformation matrix. All pixels in the front-facing image are scaled so that the distance from the stereo camera 2 is constant. This allows the difference in size of the rebar depending on the distance between the stereo camera 2 and each partial plane to be corrected in the front-facing image.

[0034] 3 is a screen diagram showing a captured image and a partial planar image of the curved surface 22A that is concave with respect to the stereo camera 2. The diagram on the left side of FIG. 3 is a screen 31A on which a stereo image of the curved surface 22A that is concave with respect to the stereo camera 2 is displayed. As shown in the screen 31A, the reinforcing bar 6 that is outside the detection range A shown in FIG. 2B is captured large on the curved surface 22A, as shown in the portion DA surrounded by a dashed line. For this reason, even if the entire image displayed on the screen 31A is converted into a frontal image as a single planar image, a part of the reinforcing bar 6 in the portion DA (the reinforcing bar 6 that is outside the detection range A) is not scaled so that the distance from the stereo camera 2 is constant.

[0035] In response to this, partial plane setting unit 12 sets a partial plane for portion DA on curved surface 22A. Image conversion unit 13 converts the image of the partial plane set for portion DA into a front-facing image, rather than converting the entire stereo image displayed on screen 31A into a front-facing image. The diagram on the right side of Fig. 3 is a screen 31B that displays the image of the partial plane set in the portion DA converted into a front-facing image. In this way, the bar arrangement inspection device 1 can calculate the front-facing image of the partial plane scaled so that the distance from the stereo camera 2 is constant by setting a partial plane in the portion DA that is strongly affected by the curvature of the curved surface 22A and determining the optimal front-facing transformation matrix for each partial plane.

[0036] In screen 31B, the vertical and horizontal reinforcing bars 6 are perpendicular to each other and are scaled so that the distance from the stereo camera 2 is constant. This allows the reinforcing bar detection unit 14 to detect the reinforcing bars 6 included in the partial plane from the image displayed on screen 31B. The detected horizontal reinforcing bars 6 are identified by markers 311A, for example, as shown in screen 31B, and the detected vertical reinforcing bars 6 are identified by markers 311B. By referring to this screen 31B, the worker can accurately recognize the vertical and horizontal reinforcing bars 6 included in the partial plane.

[0037] Although an example has been shown in which a concave curved surface relative to the stereo camera 2 is the inspection target, the bar arrangement inspection device 1 can also inspect a convex curved surface relative to the stereo camera 2. FIG. 4 is a schematic diagram showing an outline of the partial plane D4 set on the curved surface 22B that is convex with respect to the stereo camera 2, and shows the stereo camera 2 and the reinforcing bars that are the subject of its imaging, viewed from the side. In the image captured by the stereo camera 2, the reinforcing bars are captured at a virtual position P. As shown in FIG. 4, in the partial plane D4 of the curved surface 22B, the reinforcing bars 6 are positioned so as to move away from the stereo camera 2 as they move toward the end edge, so that the reinforcing bars 6 appear small in the image captured by the stereo camera 2. For this reason, even if the entire stereo image of the curved surface 22B is converted into a frontal image as an image of one plane, some of the reinforcing bars 6 assembled in the partial plane D4 (the reinforcing bars 6 outside the detection range A shown in FIG. 2B) are not scaled so that the distance between them and the stereo camera 2 is constant.

[0038] 5 is a screen diagram showing a captured image and a partial planar image of the curved surface 22B of the reinforcement that is convex with respect to the stereo camera 2. The diagram on the left side of FIG. 5 is a screen 31C on which a stereo image of the curved surface 22B that is convex with respect to the stereo camera 2 is displayed. As shown in the screen 31C, on the curved surface 22B, the reinforcing bars 6 that are outside the detection range A shown in FIG. 2B appear small, as shown in the partial DB surrounded by a dashed line. For this reason, even if the entire image displayed on the screen 31C is converted into a frontal image as a single planar image, some of the reinforcing bars 6 in the partial DB (the reinforcing bars 6 that are outside the detection range A) are not scaled so that the distance to the stereo camera 2 is constant.

[0039] In response to this, partial plane setting unit 12 sets a partial plane for partial DB on curved surface 22B. Image conversion unit 13 converts the image of the partial plane set in partial DB into a front-facing image, rather than converting the entire stereo image displayed on screen 31C into a front-facing image. The diagram on the right side of Fig. 5 is a screen 31D that displays the partial plane image set in the partial DB converted into a front-facing image. In this way, the bar arrangement inspection device 1 can calculate the front-facing image of the partial plane scaled so that the distance from the stereo camera 2 is constant by setting the partial plane in the partial DB that is strongly affected by the curvature of the curved surface 22B and determining the optimal front-facing transformation matrix for each partial plane.

[0040] In screen 31D, the vertical and horizontal reinforcing bars 6 are perpendicular to each other and are scaled so that the distance from the stereo camera 2 is constant. This allows the reinforcing bar detection unit 14 to detect the reinforcing bars 6 included in the partial plane from the image displayed on screen 31D. The detected horizontal reinforcing bars 6 are identified by marker 311A ​​as shown in screen 31D, for example, and the detected vertical reinforcing bars 6 are identified by marker 311B. By referring to this screen 31D, the worker can accurately recognize the vertical and horizontal reinforcing bars 6 included in the partial plane.

[0041] The reinforcing bar detection unit 14 uses a front-facing image of a partial plane to detect the position of the reinforcing bar included in the partial plane. For example, the reinforcing bar detection unit 14 binarizes the front-facing image by threshold processing, generates a masked image in which the binarized image is masked except for the image portion of the reinforcing bar, detects the position of the reinforcing bar in the front-facing image by counting the pixels of the image portion of the reinforcing bar, and generates reinforcing bar detection information indicating the position of the reinforcing bar.

[0042] For example, the reinforcing bar detection unit 14 uses the left and right viewpoint images of the captured image of the partial plane to calculate the amount of shift in the image of the reinforcing bar at the same position as the left-right parallax, identifies the image portion of the reinforcing bar using the parallax, and generates a mask image by setting the pixel value of the identified image portion to "1" and the other image portions to "0". The reinforcing bar detection unit 14 then detects the position in the mask image where the count number of white pixels is equal to or greater than a threshold as the position of the reinforcing bar. The reinforcing bar detection unit 14 can detect the position of the reinforcing bar in the X direction and the Y direction by rotating the mask image and counting the number of white pixels. That is, the positions of the reinforcing bars lined up vertically and the positions of the reinforcing bars lined up horizontally in the front-facing image are detected.

[0043] Next, the reinforcing bar detection unit 14 identifies a line segment in the mask image that corresponds to the reinforcing bar. If the mask image includes white pixel regions (A), (B), and (C), it is unclear which region corresponds to the image of the reinforcing bar. Therefore, the reinforcing bar detection unit 14 calculates the length D(1) of region (A), the length D(2) of region (B), and the length D(3) of region (C) on the coordinate axis passing through the white pixel regions (A), (B), and (C), the distance D(4) between region (A) and region (B), and the distance D(5) between region (B) and region (C).

[0044] The reinforcing bar detection unit 14 determines that areas between which the interval D(4) and the interval D(5) are equal to or greater than the threshold are not reinforcing bars, and determines that areas between which the interval is less than the threshold are image areas corresponding to the same reinforcing bar. For example, the reinforcing bar detection unit 14 determines that an image area indicated by a line segment having a length of D(6) that passes through area (A) and area (B) with an interval D(4) less than the threshold is an image area corresponding to the same reinforcing bar. Furthermore, since D(5) is equal to or greater than the threshold, the reinforcing bar detection unit 14 determines that there is no image area corresponding to a reinforcing bar between area (B) and area (C) in the mask image.

[0045] When the reinforcing bar detection unit 14 extracts the image area of ​​the reinforcing bars in the front-on image, it sets detection lines on each of the reinforcing bars that cross vertically and horizontally in the image area, and sets the central pixel position G of a specified pixel range area centered on the point where these detection lines cross each other as the intersection point where the reinforcing bars cross vertically and horizontally. Fig. 6 is a schematic diagram showing an outline of the process for determining the reference position of the reinforcing bars 6. The diagram on the left side of Fig. 6 shows reinforcing bars 6 that cross vertically and horizontally in a front-view image, and the diagram on the right side of Fig. 6 is a partial enlarged diagram of a pixel area F in the diagram on the left side of Fig. 6.

[0046] For example, when the reinforcing bar detection unit 14 extracts an image area of ​​reinforcing bars 6 arranged in a lattice pattern as shown in the left diagram of Fig. 6, it sets detection lines L1 and L2 on the reinforcing bars 6 that cross vertically and horizontally in the extracted image area. The detection lines are, for example, as shown in the right diagram of Fig. 6, straight lines that pass through the center position (center pixel position) of the reinforcing bar 6 in the width direction and through a pixel row made up of pixels lined up in the longitudinal direction of the reinforcing bar 6.

[0047] Next, the reinforcing bar detection unit 14 determines the central pixel position G of pixel area F, which is within a range of P1 pixels × P2 pixels and has a center at the point where the detection lines L1 and L2 intersect at right angles, as an intersection point where the reinforcing bars 6 intersect vertically and horizontally. For example, as shown in the diagram on the right side of Fig. 6, the central pixel position G is the point where the detection lines L1 and L2 intersect at right angles in a rectangular area of ​​P1 = 5 pixels × P2 = 5 pixels. This allows the intersections of the vertical and horizontal reinforcing bars 6 to be determined pixel by pixel, enabling the bar arrangement inspection device 1 to accurately detect the positions of the reinforcing bars 6 from the front-facing image. The reinforcing bar detection unit 14 adds reinforcing bar detection information indicating the positions of the portions of the mask image that have been determined to be image areas of the reinforcing bars 6 to the front-facing image, and outputs this front-facing image to the association unit 15.

[0048] The matching unit 15 matches the same reinforcing bars 6 detected in the frontal images of various partial planes. For example, the matching unit 15 matches the vertical reinforcing bars 6 detected from the frontal images of multiple partial planes set along a curved surface, using the reinforcing bar position information of the vertical reinforcing bars 6 generated by the reinforcing bar detection unit 14, with the vertical coordinates in the frontal images being the same or nearly the same (for example, position coordinates including an allowable error) in the frontal images. As a method for matching the same reinforcing bars 6, it is possible to extract image areas of the reinforcing bars 6 in the frontal images of multiple partial planes, and match partial images including these image areas as the same reinforcing bars 6. The association unit 15 outputs to the measurement processing unit 16 the partial images including the same reinforcing bar 6 extracted from each of the frontal images of the partial planes, and the reinforcing bar position information thereof.

[0049] The measurement processing unit 16 measures the reinforcing bars 6 on the curved surface based on the reinforcing bar position information of the reinforcing bars 6 associated as the same. For example, the measurement processing unit 16 determines the positions of the detection lines L1 and L2, which are straight lines passing through the intersection shown in the diagram on the right side of Fig. 6, as the reference position of the reinforcing bars 6, extracts a partial image of the reinforcing bars from the frontal image based on the reference position of the reinforcing bars 6, and measures at least one of the reinforcing bar diameter or the reinforcing bar spacing using the extracted partial image of the reinforcing bars. This enables the reinforcing bar inspection device 1 to accurately measure at least one of the reinforcing bar diameter or the reinforcing bar spacing.

[0050] The reinforcing bar inspection device 1 sequentially registers the inspection result information of the reinforcing bars 6 on the curved surface of the reinforcing bar arrangement to be inspected in the inspection result database stored in the storage unit 3 (or the memory 104 shown in FIG. 8B). The inspection result information includes not only the inspection results such as the reinforcing bar diameter, reinforcing bar spacing, and number of reinforcing bars, but also the images used for these measurements.

[0051] When a new image of reinforcing bars is taken, the measurement processing unit 16 performs correlation calculations with already registered inspection result information, and newly registers the inspection result information obtained from the photographed image that is evaluated to have a high correlation with the already registered inspection result information in the inspection result database. For example, assume that the evaluation position is a position where the distance to the reinforcing bars is 1.5 m. The evaluation position is a position where the correlation between the already registered results and the newly inspected results is evaluated, and the distance to the reinforcing bars to be inspected is a predetermined value (for example, 1.5 m). When the measurement processing unit 16 inspects the reinforcing bars 6 shown in the frontal image of the partial plane, it generates an image by reducing or enlarging this image to the image size when it was photographed at the evaluation position, and registers the generated image in the inspection result database.

[0052] The measurement processing unit 16 sets a value of 0 to pixels within a 10% length range from the central pixel position G shown in the diagram on the right side of Figure 6, of the length of the reinforcing bar 6 including the reference position, and removes it from the inspection target, and the remaining portion within the remaining 80% length range is the segment of the reinforcing bar 6 to be inspected. If the number of pixels in the segment of the reinforcing bar 6 being inspected does not reach a specified number of pixels (e.g., 512 pixels), the measurement processing unit 16 determines that the reinforcing bar 6 has not been clearly photographed, sets the pixels in that segment to a value of 0, and removes it from the inspection target.

[0053] Meanwhile, the measurement processing unit 16 acquires a scanning line consisting of the luminance values ​​of a plurality of pixels along the longitudinal direction of the reinforcing bar 6 in the frontal image for each pixel in a direction perpendicular to the longitudinal direction of the reinforcing bar 6. Next, the measurement processing unit 16 calculates a plurality of frequency conversion spectra by frequency converting (for example, FFT; Fast Fourier Transform) the luminance value of the scanning line for each pixel.

[0054] For example, the measurement processing unit 16 identifies a scanning line i corresponding to the boundary position between the reinforcing bar 6 and the background from among the multiple frequency conversion spectra, and calculates the interval between the scanning lines i. Next, the measurement processing unit 16 calculates the reinforcing bar diameter by converting the interval between the scanning lines i into a length in real space. For example, when the stereo camera 2 faces the partial plane at a shooting distance of 1.5 m, if the captured image is 200 opx in the horizontal direction, it is 0.75 mm / px in real space.

[0055] The measurement processing unit 16 performs similar processing on other frontal images associated with the frontal image of the reinforcing bar 6 for which the reinforcing bar diameter has been calculated, to calculate the reinforcing bar diameter. For example, the measurement processing unit 16 may determine the maximum or minimum value of these reinforcing bar diameters as the final reinforcing bar diameter, or may determine the average value of these reinforcing bar diameters as the final reinforcing bar diameter.

[0056] The measurement processing unit 16 may also calculate the interval between adjacent reinforcing bars 6 based on the reinforcing bar position information of the reinforcing bars 6. In this case as well, the measurement processing unit 16 performs similar processing on other frontal images associated with the frontal image of the reinforcing bar 6 for which the reinforcing bar interval has been calculated, to calculate the reinforcing bar interval. For example, the measurement processing unit 16 may take the maximum or minimum value of these reinforcing bar intervals as the final reinforcing bar interval, or may take the average value of these reinforcing bar intervals as the final reinforcing bar interval.

[0057] The start point where scanning of the image area of ​​the rebar 6 begins and the end point where scanning ends are determined by, for example, setting a scanning range that is the maximum rebar diameter in the JIS standard plus a certain margin, and setting the rebar image within the scanning range. As a result, one side of the scanning range becomes the start point and the other side becomes the end point. The start point where scanning starts and the end point where scanning ends may be set by the user using the operation unit 4.

[0058] Furthermore, scanning lines i=0, 1, 2, and 3 along the longitudinal direction of the rebar 6 are set in a direction perpendicular to the longitudinal direction of the rebar 6 (the width direction of the rebar). When the amount of change in luminance value in the image area of ​​the rebar 6 is not uniform, scanning lines i=0, 1, 2, and 3, for example, will be in a situation where the pixel row along scanning line i=3 has a high overall luminance value but the amount of change in luminance value along the scanning line is small. In this case, the pixel row along scanning line i=0 is in the shade of the sunlight, so the overall luminance value is low but the amount of change in luminance value along the scanning line is large.

[0059] Therefore, the measurement processing unit 16 performs frequency conversion on the luminance values ​​of pixel rows along each of the scanning lines i=0, 1, 2, and 3, and calculates the frequency conversion spectrum of each scanning line. For example, the frequency conversion spectra of the scanning lines i=0, 1, 2, and 3 are SP0, SP1, SP2, and SP3. When the amount of change in luminance values ​​along the scanning line i=0 is large, the frequency conversion spectrum SP0 is a spectrum with a large amplitude. On the other hand, when the amount of change in luminance values ​​along the scanning line i=3 is small, the frequency conversion spectrum SP3 is a spectrum with a small amplitude.

[0060] The measurement processing unit 16 calculates the median of the amount of change in luminance value for each frequency of the frequency conversion spectra SP0, SP1, SP2, and SP3 as the feature amount of the reinforcing bar 6. In other words, the frequency conversion spectra SP0, SP1, SP2, and SP3 are combined into a frequency conversion spectrum consisting of the median of the amount of change in luminance value for each frequency. Furthermore, the measurement processing unit 16 may synthesize the frequency transform spectrum after synthesis in each segment into a frequency transform spectrum consisting of the median value of the amount of change in luminance value for each frequency.

[0061] The measurement processing unit 16 calculates a correlation coefficient between the median of the change in luminance value at the peak frequency in the frequency conversion spectrum (the frequency conversion spectrum after the above-mentioned synthesis) obtained from the front-facing image of the captured image of the partial plane, and the median of the change in luminance value at the peak frequency in the frequency conversion spectrum obtained from the image of the reinforcing bar 6 registered in advance. Then, the measurement processing unit 16 calculates the correlation evaluation value by adding the median of the change in brightness value at the frequency of the peak weighted according to the value of the correlation coefficient, and newly registers the frequency conversion spectrum of the reinforcing bar 6 whose correlation evaluation value exceeds the threshold value.

[0062] For example, the measurement processing unit 16 performs a correlation calculation (1) that evaluates the correlation between frequency conversion spectra over the entire frequency range, a correlation calculation (2) that evaluates the correlation between the first peaks (Peak A and Peak B) that appear first in the frequency conversion spectra, and a correlation calculation (3) that evaluates the correlation between the second peaks that appear next in the frequency conversion spectra. In the correlation calculation (1), the measurement processing unit 16 calculates a correlation score (1) indicating the correlation between both spectra in a correlation evaluation range expressed as (min(PeakA,PeakB)-4) to 256. PeakA is the intensity of the first peak that first appears in a pre-registered frequency conversion spectrum, and PeakB is the intensity of the first peak that first appears in a frequency conversion spectrum obtained in the current examination. In the correlation calculation (2), the measurement processing unit 16 calculates a correlation score (2) indicating the correlation between both spectra within a correlation evaluation range represented by (PeakA)±10. In the correlation calculation (3), the measurement processing unit 16 calculates a correlation score (3) indicating the correlation between both spectra within a correlation evaluation range represented by (PeakA×2)±10.

[0063] The measurement processing unit 16 determines weighting coefficients according to the respective values ​​of the correlation score (1), correlation score (2), and correlation score (3). Then, the measurement processing unit 16 calculates a total correlation score, for example, according to the following formula. If the total correlation score is equal to or greater than a threshold, the measurement processing unit 16 determines that the inspection result information using the frequency conversion spectrum obtained in the current inspection is similar to previously registered inspection result information, and registers the inspection result information in the inspection result database. This allows the reinforcement inspection device 1 to newly register highly reliable inspection results that are highly correlated with already registered inspection results. Total correlation score = (correlation score (1) x 0.5) + (correlation score (2) x 0.2) + (correlation score (3) x 0.3)

[0064] The measurement processing unit 16 may also divide the image of the reinforcing bar 6 extracted from the frontal image into a plurality of partial images, and infer the reinforcing bar diameter using a learning model that outputs the diameter of the reinforcing bar 6 when the plurality of partial images are input. For example, the measurement processing unit 16 receives input of partial images obtained by dividing an image of a single reinforcing bar 6 into a plurality of partial images, and infers the reinforcing bar diameter using a learning model that infers the reinforcing bar diameter as reinforcing bar inspection information. This learning model is generated for each frontal image of a partial plane. This enables the reinforcing bar inspection device 1 to accurately measure the reinforcing bar diameter.

[0065] A learning device (not shown in FIG. 1) that generates a learning model generates the learning model by learning to infer rebar diameter from learning data. The learning data is a data set including multiple partial image data of rebar 6 and correct answer labels, which are input data for the learning model. The correct answer label is information indicating the rebar diameter corresponding to each of the multiple partial image data. The learning model is stored in the memory unit 3. For example, deep learning, neural network, genetic programming, functional logic programming, or support vector machine is used as the learning algorithm.

[0066] The measurement processing unit 16 extracts an image of the reinforcing bar 6 along the longitudinal direction from the front-facing image of the partial plane. Furthermore, the measurement processing unit 16 sequentially extracts partial images of the same size from the front-facing image of the partial plane. These partial images are, for example, square images with the same number of pixels in both length and width, and the reinforcing bar diameter is R1. The measurement processing unit 16 inputs the multiple partial images into a learning model to perform inference, and calculates, for each partial image, the probability that a reinforcing bar 6 with a reinforcing bar diameter R1 is captured and the probability that something other than the reinforcing bar 6 is captured.

[0067] There are 16 types of rebars 6 based on the JIS standard, for example. The 16 types of rebars 6 are given "names" such as D4, D5, D6, D8, D10, D13, D16, D19, D22, D25, D29, D32, D35, D38, D41, and D51. The names indicate the diameter of the rebar rounded off from the nominal diameter.

[0068] The nominal diameter of a rebar designated D10 is 9.53 mm, that of a rebar designated D13 is 12.7 mm, and that of a rebar designated D16 is 15.9 mm. Note that rebars generally used as the framework of buildings are of D10 or higher.

[0069] Next, the learning model averages the inference result data for all partial image data extracted from the same rebar image. For example, the inference result data 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 processing unit 16 determines the rebar diameter R1 based on the inference result data. In this way, the measurement processing unit 16 can accurately detect the rebar diameter R1 by using the learning model, and can determine the type of rebar based on the rebar diameter R1. In addition, the learning model may be one that outputs the distance between adjacent rebars 6 when multiple partial images of a rebar 6 are input, or one that infers the distance between nodes in a single rebar 6.

[0070] FIG. 7 is a flowchart showing the bar arrangement inspection method according to the first embodiment. The three-dimensional image data acquisition unit 11 generates three-dimensional image data of a curved surface in the reinforcement by using the stereo image of the reinforcement (step ST1). The partial plane setting unit 12 uses the three-dimensional image data to set a plurality of partial planes along the curved surface of the reinforcement shown in the stereo image (step ST2). The image conversion unit 13 converts the image of the partial plane into a frontal image in which the reinforcing bars are perpendicular to each other (step ST3). The reinforcing bar detection unit 14 detects the positions of the reinforcing bars included in the partial plane by using the frontal image (step ST4). The association unit 15 associates the same reinforcing bars among those detected in the multiple frontal images (step ST5). The measurement processing unit 16 measures the reinforcing bars on the curved surface based on the position information of the reinforcing bars associated as identical bars (step ST6). By carrying out this method, the bar arrangement inspection device 1 can measure the reinforcing bars on a curved surface that are made up of reinforcing bars arranged in a lattice pattern.

[0071] Next, a hardware configuration for realizing the functions of the bar arrangement inspection device 1 will be described. The functions of the three-dimensional image data acquisition unit 11, partial plane setting unit 12, image conversion unit 13, reinforcing bar detection unit 14, association unit 15, and measurement processing unit 16 included in the reinforcing bar inspection device 1 are realized by a processing circuit. That is, the reinforcing bar inspection device 1 includes a processing circuit for executing each process from step ST1 to step ST6 shown in Fig. 7. The processing circuit may be dedicated hardware, or may be a CPU (Central Processing Unit) that executes a program stored in a memory.

[0072] Fig. 8A is a block diagram showing a hardware configuration for realizing the functions of the bar arrangement inspection device 1. Fig. 8B is a block diagram showing a hardware configuration for executing software for realizing the functions of the bar arrangement inspection device 1. In Figs. 8A and 8B, the input interface 100 is an interface that relays image information output from the stereo camera 2 to the bar arrangement inspection device 1, information output from the storage unit 3 to the bar arrangement inspection device 1, and operation information output from the operation unit 4 to the bar arrangement inspection device 1. The output interface 101 is an interface that relays the bar arrangement inspection results output from the bar arrangement inspection device 1 to the display unit 5.

[0073] When the processing circuit is the dedicated hardware processing circuit 102 shown in FIG. 8A, 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 three-dimensional image data acquisition unit 11, partial plane setting unit 12, image conversion unit 13, reinforcing bar detection unit 14, matching unit 15, and measurement processing unit 16 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.

[0074] When the processing circuit is the processor 103 shown in Fig. 8B, the functions of the three-dimensional image data acquisition unit 11, partial plane setting unit 12, image conversion unit 13, reinforcing bar detection unit 14, association unit 15 and measurement processing unit 16 included in the reinforcing bar 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. The memory 104 is, for example, the storage unit 3 shown in Fig. 1.

[0075] The processor 103 reads out and executes the program of the reinforcement inspection application stored in the memory 104, thereby realizing the functions of the three-dimensional image data acquisition unit 11, partial plane setting unit 12, image conversion unit 13, reinforcing bar detection unit 14, matching unit 15 and measurement processing unit 16 provided in the reinforcement inspection device 1. For example, the reinforcing bar inspection device 1 includes a memory 104 for storing the above-mentioned program, which, when executed by the processor 103, results in the processing of steps ST1 to ST6 shown in Fig. 7 being executed. This program causes a computer to execute the procedures or methods of processing performed by the three-dimensional image data acquisition unit 11, the partial plane setting unit 12, the image conversion unit 13, the reinforcing bar detection unit 14, the association unit 15, and the measurement processing unit 16. The memory 104 may be a computer-readable storage medium in which a program for causing a computer to function as the three-dimensional image data acquisition unit 11, the partial plane setting unit 12, the image conversion unit 13, the reinforcing bar detection unit 14, the association unit 15, and the measurement processing unit 16 is stored.

[0076] Memory 104 may be, for example, a non-volatile or volatile semiconductor memory such as a Random Access Memory (RAM), a Read Only Memory (ROM), a flash memory, an Erasable Programmable Read Only Memory (EPROM), an Electrically-EPROM (EEPROM), a magnetic disk, a flexible disk, an optical disk, a compact disk, a mini disk, a DVD, etc.

[0077] Some of the functions of the three-dimensional image data acquisition unit 11, partial plane setting unit 12, image conversion unit 13, rebar detection unit 14, association unit 15, and measurement processing unit 16 included in the reinforcing bar inspection device 1 may be realized by dedicated hardware, and the remaining functions may be realized by software or firmware. For example, the three-dimensional image data acquisition unit 11 has its functions realized by a processing circuit 102, which is dedicated hardware, and the partial plane setting unit 12, image conversion unit 13, rebar detection unit 14, association unit 15, and measurement processing unit 16 have their functions realized by a processor 103 reading and executing programs stored in a memory 104. Thus, the processing circuitry may implement the above-described functions through hardware, software, firmware or a combination of these.

[0078] As described above, the reinforcing bar inspection device 1 according to the first embodiment includes a three-dimensional image data acquisition unit 11 that generates three-dimensional image data of a curved surface using stereo images of reinforcing bars, a partial plane setting unit 12 that sets multiple partial planes along the curved surface in the stereo images, an image conversion unit 13 that converts the partial plane images into frontal images in which the reinforcing bars 6 are perpendicular to each other, a reinforcing bar detection unit 14 that detects the positions of the reinforcing bars 6 included in the partial planes using the frontal images, a matching unit 15 that matches the same reinforcing bars 6 detected in the multiple frontal images, and a measurement processing unit 16 that measures the reinforcing bars 6 on the curved surface based on the reinforcing bar position information of the reinforcing bars 6 that are matched as the same. As a result, the reinforcing bar inspection device 1 can measure the reinforcing bars 6 on a curved surface consisting of reinforcing bars 6 arranged in a lattice pattern.

[0079] In the reinforcing bar inspection device 1 according to the first embodiment, the reinforcing bar detection unit 14 sets detection lines L1 and L2 on the reinforcing bars 6 that cross vertically and horizontally in the front-facing image, and sets the central pixel position G of an area of ​​a predetermined pixel range centered on the point where the detection lines L1 and L2 cross as the intersection point where the reinforcing bars 6 cross vertically and horizontally. The measurement processing unit 16 determines the position of the straight line passing through the intersection point as the reference position of the reinforcing bars 6, and measures at least one of the reinforcing bar diameter or the reinforcing bar spacing based on the determined reference position of the reinforcing bars 6. This allows the reinforcing bar inspection device 1 to accurately detect the position of the reinforcing bars 6 from the front-facing image, and to accurately measure the reinforcing bars 6.

[0080] In the bar arrangement inspection device 1 according to the first embodiment, the measurement processing unit 16 extracts a partial image of the reinforcing bar 6 from the frontal image based on the reference position of the reinforcing bar 6, and measures the reinforcing bar diameter using the extracted partial image of the reinforcing bar 6. This enables the bar arrangement inspection device 1 to accurately measure the reinforcing bar diameter.

[0081] In the bar arrangement inspection device 1 according to the first embodiment, the measurement processing unit 16 estimates the reinforcing bar diameter using a machine learning model that outputs the reinforcing bar diameter when a partial image of the reinforcing bar 6 is input. This enables the bar arrangement inspection device 1 to accurately measure the reinforcing bar diameter.

[0082] The reinforcing bar inspection method according to the first embodiment includes step ST1 in which the three-dimensional image data acquisition unit 11 generates three-dimensional image data of the curved surface using stereo images of the reinforcing bars, step ST2 in which the partial plane setting unit 12 sets a plurality of partial planes along the curved surface, step ST3 in which the image conversion unit 13 converts the image of the partial plane into a frontal image in which the reinforcing bars 6 are perpendicular to each other, step ST4 in which the reinforcing bar detection unit 14 detects the position of the reinforcing bars 6 included in the partial plane using the frontal image, step ST5 in which the association unit 15 associates the same reinforcing bars 6 detected in the plurality of frontal images, and step ST6 in which the measurement processing unit 16 measures the reinforcing bars 6 on the curved surface based on the reinforcing bar position information of the reinforcing bars 6 associated as the same. The reinforcing bar inspection device that executes this method is capable of measuring the reinforcing bars 6 on a curved surface consisting of reinforcing bars 6 arranged in a lattice pattern.

[0083] A computer that executes the program according to the first embodiment functions as a reinforcing bar inspection device 1 that includes a three-dimensional image data acquisition unit 11 that generates three-dimensional image data of a curved surface using stereo images of reinforcing bars, a partial plane setting unit 12 that sets multiple partial planes along the curved surface in the stereo images, an image conversion unit 13 that converts the partial plane images into frontal images in which reinforcing bars 6 are perpendicular to each other, a reinforcing bar detection unit 14 that detects the positions of reinforcing bars 6 included in the partial planes using the frontal images, a matching unit 15 that matches identical reinforcing bars 6 detected in the multiple frontal images, and a measurement processing unit 16 that measures reinforcing bars 6 on the curved surface based on reinforcing bar position information of reinforcing bars 6 that are matched as identical. This makes it possible to provide a reinforcing bar inspection device 1 that measures reinforcing bars 6 on a curved surface consisting of reinforcing bars 6 arranged in a lattice pattern.

[0084] It is possible to combine the embodiments, modify any of the components of the embodiments, or omit any of the components of the embodiments. [Explanation of symbols]

[0085] 1 Reinforcement inspection device, 2 stereo camera, 2A shooting field of view, 3 memory unit, 4 operation unit, 5 display unit, 6 reinforcing bar, 11 three-dimensional image data acquisition unit, 12 partial plane setting unit, 13 image conversion unit, 14 reinforcing bar detection unit, 15 matching unit, 16 measurement processing unit, 21 plane, 22A, 22B curved surface, 31A, 31B, 31C, 31D screen, 100 input interface, 101 output interface, 102 processing circuit, 103 processor, 104 memory, 311A, 311B markers.

Claims

1. A reinforcing bar inspection device for inspecting a reinforcing bar having a curved surface made of reinforcing bars arranged in a lattice pattern, a three-dimensional image data acquisition unit that generates three-dimensional image data of the curved surface using the stereo image of the reinforcement; a partial plane setting unit that sets a plurality of partial planes along the curved surface in the stereo image; An image conversion unit that converts the partial plane image into a frontal image in which the reinforcing bars are perpendicular to each other; A reinforcing bar detection unit that detects the position of a reinforcing bar included in the partial plane by using the frontal image; A matching unit that matches identical reinforcing bars detected in the plurality of frontal images; A measurement processing unit that measures the reinforcing bars on the curved surface based on reinforcing bar position information of the reinforcing bars associated as the same reinforcing bar. A reinforcing bar inspection device characterized by the above.

2. The reinforcing bar detection unit sets detection lines on the reinforcing bars that cross vertically and horizontally in the frontal image, and determines a central pixel position of an area of ​​a predetermined pixel range centered on a point where the detection lines cross as an intersection point where the reinforcing bars cross vertically and horizontally, The measurement processing unit determines the position of a straight line passing through the intersection as a reference position of the reinforcing bar, and measures at least one of a reinforcing bar diameter or a reinforcing bar interval based on the determined reference position of the reinforcing bar.

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

3. The measurement processing unit extracts a partial image of the reinforcing bar from the front image based on a reference position of the reinforcing bar, and measures a diameter of the reinforcing bar using the extracted partial image of the reinforcing bar.

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

4. The measurement processing unit estimates the diameter of the reinforcing bar using a machine learning model that outputs the diameter of the reinforcing bar when a partial image of the reinforcing bar is input.

4. The reinforcing bar inspection device according to claim 3.

5. A method for inspecting a reinforcing bar having a curved surface made of reinforcing bars arranged in a lattice pattern using a reinforcing bar inspection device, comprising: A three-dimensional image data acquisition unit generates three-dimensional image data of the curved surface using the stereo image of the reinforcement; A partial plane setting unit sets a plurality of partial planes along the curved surface in the stereo image; An image conversion unit converts the image of the partial plane into a frontal image in which the reinforcing bars are perpendicular to each other; A step in which a reinforcing bar detection unit detects a position of a reinforcing bar included in the partial plane by using the frontal image; A step in which an association unit associates identical reinforcing bars detected in the plurality of frontal images; The measurement processing unit measures the reinforcing bars on the curved surface based on reinforcing bar position information of the reinforcing bars associated as the same reinforcing bar. A reinforcement inspection method comprising:

6. Computer, A reinforcing bar inspection device for inspecting a reinforcing bar having a curved surface made of reinforcing bars arranged in a lattice pattern, a three-dimensional image data acquisition unit that generates three-dimensional image data of the curved surface using the stereo image of the reinforcement; a partial plane setting unit that sets a plurality of partial planes along the curved surface in the stereo image; An image conversion unit that converts the partial plane image into a frontal image in which the reinforcing bars are perpendicular to each other; A reinforcing bar detection unit that detects the position of a reinforcing bar included in the partial plane by using the frontal image; A matching unit that matches identical reinforcing bars detected in the plurality of frontal images; a measurement processing unit that measures the reinforcing bars on the curved surface based on reinforcing bar position information of the reinforcing bars associated as identical reinforcing bars; A program that functions as a

Citation Information

Patent Citations

  • Bar arrangement inspection device, bar arrangement inspection method and program

    JP2022164949A