A method for automatic analysis of spacing between steel bars in low-light environment

Abstract

This application provides an automatic analysis method for rebar spacing in low-light environments, comprising: deploying a supplementary lighting unit and a high-sensitivity camera at a low-light construction site to collect supplementary lighting on the rebar arrangement area to obtain an original image containing the complete rebar arrangement; extracting the edge width of each rebar along its length at multiple cross-sections based on the outline range of each rebar, and taking the average of the edge widths of each cross-section as the measured diameter of the rebar; extracting the centerline position of each rebar based on its outline edge and measured diameter, and determining the pairwise combination of adjacent rebars according to the rebar arrangement order; extracting the distance between the centerlines of adjacent rebars as the center spacing based on their centerline positions, and simultaneously subtracting the radii on both sides based on the measured diameter to obtain the net spacing between the surfaces of adjacent rebars.

Description

Technical Field

[0001] This invention relates to the field of information technology, and in particular to an automatic analysis method for steel bar spacing under low light conditions. Background Technology

[0002] In the fields of civil engineering inspection and building safety assessment, accurate measurement of rebar spacing is crucial for ensuring the load-bearing capacity and durability of concrete structures. Especially in low-light environments at construction sites, traditional manual inspection methods are inefficient and susceptible to subjective biases. Therefore, developing automated visual analysis methods is of significant practical importance. Existing automated image processing methods typically rely on edge detection technology, locating rebar boundaries by calculating the intensity of pixel grayscale changes in an image. However, under insufficient lighting conditions, image noise increases and boundaries become blurred. Directly applying these methods leads to numerous irregular, jagged fluctuations in the detected boundary lines. These fluctuations are not the true geometric contours of the rebar but artifacts caused by imaging noise and uneven lighting, severely interfering with the accuracy of spacing calculations. To suppress jagged artifacts, the conventional approach is to enhance the smoothness constraint on the image spatial gradient, i.e., increase the weight of spatial gradient information in the algorithm. This forces the detected boundary lines to be straighter and smoother, thereby reducing jagged density and protrusion, and improving the stability of coordinate fitting and spacing measurement. However, the actual surface of rebar is not perfectly smooth; its threaded ribs create periodic undulating contours, which are inherent geometric features of the rebar. If the spatial gradient weights are excessively increased, the algorithm will be unable to distinguish between jagged artifacts caused by noise and the true boundary undulations caused by thread ribs. It will treat both as noise that needs smoothing, resulting in the incorrect smoothing of thread features. In practical applications, when the system mistakenly treats the periodic unevenness of the thread ribs as jagged artifacts for smoothing, the visual boundary of the rebar will shrink inward, causing the measured rebar diameter to be smaller than the actual value, resulting in a systematic negative bias. Since the calculation of rebar spacing depends on the positioning of adjacent rebar boundaries, the diameter deviation will further propagate and amplify the error in spacing measurement. Therefore, in low-light environments, how to effectively eliminate boundary jagged artifacts caused by imaging noise without erasing the true thread features, thereby ensuring the dual accuracy of rebar diameter and spacing measurements, becomes a key issue for the reliable application of automated inspection technology. Summary of the Invention

[0003] This invention provides an automatic analysis method for rebar spacing under low-light conditions, comprising: In low-light construction sites, supplementary lighting units and high-sensitivity cameras are deployed to collect supplementary lighting data on areas where steel bars are arranged, thereby obtaining original images containing the complete arrangement of steel bars. The original image is subjected to brightness enhancement and noise reduction processing to identify the boundary between the rebar area and the background area, extract the contour range of each rebar, and obtain the separation result of a single rebar and the number of rebars. Based on the outline range of each steel bar, the edge width of the steel bar is extracted from multiple sections along the length of the steel bar, and the average value of the edge width of each section is taken as the actual measured diameter of the steel bar. Based on the outline edge of each steel bar and the measured diameter, the centerline position of each steel bar is extracted, and the pair combination of adjacent steel bars is determined according to the steel bar arrangement order. The distance between the centerlines of adjacent steel bars is extracted as the center spacing based on the centerline positions of the adjacent steel bars. At the same time, the net spacing between the surfaces of adjacent steel bars is obtained by subtracting the radii on both sides based on the measured diameter. The net spacing is repeatedly extracted at multiple sections along the length of the reinforcing bar to evaluate whether the net spacing of the same pair of adjacent reinforcing bars is consistent at different sections. Uneven spacing areas caused by the inclination or bending of the reinforcing bars are identified, and the average value of the cross sections of the uneven areas is taken as the corrected net spacing. By integrating the net spacing and center-to-center spacing of adjacent reinforcing bars, comparing them with the allowable spacing range specified in the construction drawings, marking areas that exceed the allowable range, and generating a reinforcing bar spacing compliance analysis result.

[0004] Furthermore, the deployment of supplementary lighting units and high-sensitivity cameras at the low-light construction site to supplement lighting and collect images of the rebar arrangement area, obtaining an original image containing the complete rebar arrangement, includes: Based on the ambient light intensity at the construction site, multiple sets of supplementary lighting units are arranged above and to the side of the steel bar arrangement area. The irradiation angle and output brightness of the supplementary lighting units are adjusted so that the irradiation range covers the entire surface of the steel bar arrangement area. A high-sensitivity camera is used to acquire images of the area where the steel bars are arranged under supplementary lighting conditions. The photosensitive parameters of the camera are adapted and adjusted according to the brightness of the supplementary lighting conditions to obtain an original image containing the complete arrangement of the steel bars.

[0005] Furthermore, the process of enhancing brightness and denoising the original image, identifying the boundary between the rebar area and the background area, extracting the contour range of each rebar, and obtaining the separation result of a single rebar and the number of rebars includes: Obtain the grayscale distribution of the original image, and perform brightness enhancement processing on the original image based on the proportion of low-brightness pixels in the grayscale distribution to obtain the enhanced image; The enhanced image is denoised using Gaussian filtering to obtain a preprocessed image; Based on the grayscale distribution of the preprocessed image, the Otsu thresholding method is used to determine the boundary threshold between the steel reinforcement area and the background area, resulting in a binary segmentation image. Connected component marking is performed on the rebar region in the binary segmentation image. The contour range of each connected component is extracted as the contour boundary of a single rebar. The number of connected components is counted to obtain the number of rebars, and the separation result of a single rebar and the number of rebars are output.

[0006] Furthermore, the step of extracting the edge width of each reinforcing bar from multiple cross-sections along its length based on the outline range of each reinforcing bar, and taking the average of the edge widths of each cross-section as the measured diameter of the reinforcing bar, includes: Based on the outline range of a single steel bar, determine the starting and ending positions of the outline along the length of the steel bar, and divide multiple cross-sectional positions according to a preset interval. For each cross-section location, extract the left boundary pixel coordinates and right boundary pixel coordinates where the cross-section intersects with the steel bar outline, and calculate the number of pixels between the two as the edge width corresponding to the cross-section; Summarize the edge widths at each cross-section location, calculate the arithmetic mean of all edge widths, and use the arithmetic mean as the measured diameter of the reinforcing bar.

[0007] Furthermore, the step of extracting the centerline position of each reinforcing bar based on its outline edge and measured diameter, and determining the pairwise combination of adjacent reinforcing bars according to the bar arrangement order, includes: Based on the outline edge of a single steel bar and the measured diameter, the left boundary pixel coordinates and the right boundary pixel coordinates of the outline edge are extracted, and the arithmetic mean of the two is taken as the center point coordinates of the cross section. The center point coordinates of each cross section are connected to form the center line position of the steel bar. Based on the centerline position of each steel bar, the coordinate value of the centerline of each steel bar in the arrangement direction is extracted, and the steel bars are sorted in ascending order of the coordinate values ​​to obtain the steel bar arrangement order; Based on the arrangement order of the reinforcing bars, two adjacent reinforcing bars are marked as a pair of adjacent reinforcing bars, and the pair combinations of adjacent reinforcing bars are obtained by iterating through them in turn.

[0008] Furthermore, the step of extracting the distance between the centerlines of adjacent reinforcing bars as the center-to-center spacing based on their centerline positions, and simultaneously subtracting the radii from both sides based on the measured diameter to obtain the net spacing between the surfaces of adjacent reinforcing bars, includes: Based on the centerline position of each pair of adjacent reinforcing bars, the distance between the centerlines of the two reinforcing bars along the arrangement direction is extracted as the center-to-center spacing of the pair of adjacent reinforcing bars. Based on the center-to-center distance and the measured diameters of the two reinforcing bars, the sum of half the measured diameters of the two reinforcing bars is subtracted from the center-to-center distance to obtain the theoretical net distance calculated based on the centerline. Identify the position of the outer surface contour edge of two steel bars that are close to each other, and obtain the measured net spacing based on the gap width between the outer surface contour edges.

[0009] Furthermore, the step of identifying the outer surface contour edge positions of two reinforcing bars on their closest sides, and obtaining the net spacing between adjacent reinforcing bar surfaces based on the gap width between the outer surface contour edges, includes: Based on the centerline position of each pair of adjacent reinforcing bars and their respective measured diameters, the coordinate difference between the centerlines of the two reinforcing bars along the arrangement direction is extracted, and the absolute value of the coordinate difference is taken as the center spacing. Based on the center spacing and centerline position, determine the outer surface contour edge position of the two reinforcing bars on the side that is close to each other. The outer surface contour edge position of one reinforcing bar on the side that is close to each other is the contour boundary of the reinforcing bar towards the other reinforcing bar, and the outer surface contour edge position of the other reinforcing bar on the side that is close to each other is the contour boundary of the reinforcing bar towards the previous reinforcing bar. Extract the coordinate difference between the two outer surface contour edges in the arrangement direction, and use the absolute value of the coordinate difference as the gap width between the outer surface contour edges; The gap width is used as the net spacing between adjacent steel bar surfaces.

[0010] Furthermore, the process of repeatedly extracting the net spacing along the length of the reinforcing bars at multiple cross-sections, evaluating whether the net spacing of the same pair of adjacent reinforcing bars is consistent at different cross-sections, identifying areas of uneven spacing caused by the inclination or bending of the reinforcing bars, and taking the cross-sectional mean value of the uneven areas as the corrected net spacing includes: For the same pair of adjacent reinforcing bars, multiple cross-sectional positions are divided along the length of the reinforcing bars. The net spacing value of the pair of adjacent reinforcing bars is extracted at each cross-sectional position to form a net spacing sequence. Based on the net spacing sequence, the global mean of all net spacing values ​​is calculated as the deviation detection benchmark value; For each cross-sectional location, calculate the difference between the net spacing value and the global mean, and take the absolute value of the difference as the deviation amplitude; The deviation amplitude is compared with a preset deviation threshold. If the deviation amplitude exceeds the deviation threshold, the cross-sectional position is determined to be an uneven section. Based on the distribution of non-uniform sections along the length of the reinforcing bars, multiple consecutive non-uniform sections are marked as the same non-uniform section. For the non-uniform section, the arithmetic mean of the net spacing values ​​of each cross-section position within the section is calculated as the corrected net spacing, and the value of the corresponding section in the original net spacing sequence is replaced.

[0011] Furthermore, the process integrates the net spacing and center-to-center spacing of adjacent reinforcing bars, compares them with the allowable spacing range specified in the construction drawings, marks areas exceeding the allowable range, and generates a reinforcing bar spacing compliance analysis result, including: Integrate the net spacing and center-to-center spacing of each pair of adjacent reinforcing bars, and retrieve the allowable spacing range specified in the construction drawings; Compare the net spacing and center spacing with the allowable spacing range respectively. If they exceed the allowable spacing range, mark the pair of reinforcing bars as an over-limit area. The location information and excess values ​​of all areas exceeding the limit are summarized to generate the compliance analysis results of the rebar spacing.

[0012] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects: This invention discloses an automatic analysis method for rebar spacing in low-light environments. Addressing the challenges of difficult image acquisition of rebar arrangement at low-light construction sites, low accuracy in spacing measurement, and uneven spacing due to rebar tilting or bending, this invention proposes an integrated solution. By deploying a supplementary lighting unit and a high-sensitivity camera, combined with brightness enhancement and noise reduction processing, this invention acquires clear images of the rebar arrangement and accurately extracts the rebar outline and measured diameter, thereby determining the centerline position and the center-to-center spacing and net spacing of adjacent rebars. Particularly for rebar tilting or bending issues, this invention extracts the net spacing from multiple cross-sections along the length direction, identifies uneven areas and corrects them by averaging, and finally integrates the spacing data with construction drawings for comparison, marking areas exceeding standards and generating compliance analysis results. This invention significantly improves the automation and accuracy of rebar spacing measurement in low-light environments, effectively solves the problem of spacing analysis in complex environments, and provides reliable technical support for construction quality control. Attached Figure Description

[0013] Figure 1 This is a flowchart of an automatic analysis method for rebar spacing under low light conditions according to the present invention.

[0014] Figure 2 This is a schematic diagram of an automatic analysis method for rebar spacing under low light conditions according to the present invention.

[0015] Figure 3 This is another schematic diagram of an automatic analysis method for rebar spacing under low light conditions according to the present invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and thoroughly described below with reference to the accompanying drawings. The described embodiments are merely some embodiments of the present invention.

[0017] like Figures 1-3 This embodiment of an automatic analysis method for rebar spacing under low light conditions may specifically include: S101. Deploy supplementary lighting units and high-sensitivity cameras at the low-light construction site to collect supplementary lighting data on the area where the steel bars are arranged, and obtain the original image containing the complete arrangement of the steel bars.

[0018] Based on the ambient light intensity at the construction site, multiple sets of supplementary lighting units are arranged above and to the sides of the rebar arrangement area. The illumination angle and output brightness of the supplementary lighting units are adjusted to ensure that the illumination range covers the entire surface of the rebar arrangement area, achieving uniformly distributed supplementary lighting conditions. A high-sensitivity camera is used to acquire images of the rebar arrangement area under the supplementary lighting conditions. The camera's photosensitive parameters are adapted and adjusted according to the brightness of the supplementary lighting conditions. After acquisition, an original image containing the complete rebar arrangement is obtained.

[0019] In one implementation, the rebar arrangement area in low-light construction sites is typically located in underground foundation pits or inside building structures, where natural light is difficult to directly illuminate, resulting in low ambient light intensity. To address this scenario, supplementary lighting units are arranged above and to the sides of the rebar arrangement area. The upper supplementary lighting unit provides vertical downward main illumination, while the side supplementary lighting unit provides oblique supplementary illumination. The illumination ranges of the two sets of supplementary lighting units overlap, covering the entire surface of the rebar arrangement area, eliminating shadow areas caused by unidirectional illumination, thereby achieving uniformly distributed supplementary lighting conditions.

[0020] Specifically, the output brightness of the supplementary lighting unit is adjusted according to the area of ​​the rebar arrangement, and the illumination angle is directionally adjusted according to the rebar arrangement direction to ensure that the light is evenly spread along the length of the rebar. Based on the above supplementary lighting conditions, a high-sensitivity camera is used to acquire images of the rebar arrangement area. The camera's photosensitive parameters, including exposure time and gain value, are adapted and adjusted according to the brightness level of the supplementary lighting conditions. The exposure time and brightness level are inversely related, and the gain value is appropriately increased when the brightness is insufficient. After acquisition, an original image containing the complete rebar arrangement is obtained.

[0021] It should be noted that the measurements of rebar diameter and spacing in subsequent steps are calculated within the image in pixels. To convert the pixel-level measurements to actual millimeters, the image resolution needs to be pre-calibrated. The calibration method is as follows: place a calibration object of known size, such as a ruler of length L millimeters, in the rebar arrangement area; use a camera to capture images of the calibration object under the same shooting conditions; and measure the pixel length P of the calibration object in the image. pixels Calculate the image resolution R=P pixels / L (unit: pixels / mm). All subsequent measurements in pixels are converted to actual distance in millimeters by dividing by the image resolution R.

[0022] S102. Perform brightness enhancement and noise reduction on the original image, identify the boundary between the rebar area and the background area, extract the contour range of each rebar, and obtain the separation result of a single rebar and the number of rebars.

[0023] The grayscale distribution of the original image is obtained. Based on the proportion of low-brightness pixels in the grayscale distribution, the original image is enhanced to increase the grayscale value of dark areas, resulting in an enhanced image. Gaussian filtering is then applied to the enhanced image to denoise and suppress random noise, resulting in a denoised preprocessed image. Based on the grayscale distribution of the preprocessed image, the Otsu thresholding method is used to determine the boundary threshold between the rebar region and the background region. Pixels with grayscale values ​​below the boundary threshold are marked as rebar regions, and pixels with grayscale values ​​above the boundary threshold are marked as background regions, resulting in a binary segmentation image of the rebar region and the background region. For the rebar region in the binary segmentation image, connected components are marked along the outer boundary of the rebar region. The contour range of each connected component is extracted as the contour boundary of a single rebar. The number of connected components is counted to obtain the number of rebars. Subsequently, the completeness of the contour extraction is verified based on the number of rebars, and the separation result of a single rebar and the number of rebars are output for subsequent spacing statistics.

[0024] In one implementation, the original image exhibits an overall dark grayscale distribution, with low-brightness pixels concentrated on the surface of the reinforcing steel. To address this characteristic, brightness enhancement processing maps the grayscale values ​​of the dark areas to the mid-to-high grayscale range through linear stretching, amplifying the grayscale difference between the reinforcing steel surface and the background, thereby obtaining an enhanced image with increased brightness.

[0025] Specifically, Gaussian filtering uses a two-dimensional Gaussian kernel to perform convolution operations on the enhanced image. The window size of the Gaussian kernel is adapted to the image resolution. During the convolution process, the output value of each pixel is determined by the weighted average of its neighboring pixels. The weights decrease from the center to the edge according to the Gaussian distribution. Random noise is smoothed and suppressed during the weighted averaging process, resulting in a denoised preprocessed image.

[0026] It should be noted that the Otsu thresholding method is an adaptive threshold determination method based on the principle of maximizing inter-class variance. This method divides the gray-level histogram of the preprocessed image into foreground and background classes, traverses all possible gray-level values ​​as candidate thresholds, and calculates the inter-class variance between the foreground and background classes for each candidate threshold. The inter-class variance reflects the degree of difference in the mean gray-level values ​​of the two classes of pixels. The gray-level value that maximizes the inter-class variance is selected as the boundary threshold. In rebar images acquired under low-light conditions, the rebar surface exhibits dark features corresponding to low gray-level ranges, while the background area exhibits light features corresponding to high gray-level ranges under supplementary lighting. Therefore, pixels with gray-level values ​​below the boundary threshold are marked as rebar areas, and pixels with gray-level values ​​above the boundary threshold are marked as background areas. Based on the above binary segmented image, the classic two-pass connected component labeling algorithm is used to assign labels to the pixels in the rebar areas. The input of this algorithm is the binary segmented image, and the output is the labeled connected component image. The specific process includes a first pass of scanning line by line to assign temporary labels and record equivalence relations, and a second pass of merging labels according to equivalence relations, thereby classifying adjacent steel bar pixels into the same connected region.

[0027] In one embodiment, each independent connected region corresponds to the projection area of ​​a steel bar, and the outer boundary of the connected region is the outline boundary of the steel bar. The number of all connected regions is counted to obtain the number of steel bars in the image, and the outline range of each connected region is output as the separation result of a single steel bar.

[0028] S103. Based on the outline range of each steel bar, extract the edge width of the steel bar at multiple sections along the length of the steel bar, and take the average value of the edge width of each section as the measured diameter of the steel bar.

[0029] Based on the outline of a single rebar, the start and end positions of the outline are determined along the length of the rebar. Multiple cross-sectional positions are then divided between these positions at preset intervals, resulting in a sequence of cross-sectional positions distributed along the length of the rebar. For each cross-sectional position in this sequence, the left and right boundary pixel coordinates of the intersection between the cross-section and the rebar outline are extracted. The number of pixels between these coordinates is calculated as the edge width corresponding to that cross-sectional position. The edge widths of all cross-sectional positions are then summed to form an edge width sequence. Based on this edge width sequence, the arithmetic mean of all edge widths is calculated, and this arithmetic mean is used as the measured diameter of the rebar.

[0030] In one embodiment, the outline of a single reinforcing bar presents as a strip-shaped region extending along its length, with the starting and ending points of the outline corresponding to the two ends of the reinforcing bar in the image. Along the length of the reinforcing bar, multiple cross-sectional positions are evenly divided between the starting and ending points at preset intervals, forming a sequence of cross-sectional positions. The preset interval is determined as follows: the rib spacing of threaded reinforcing bars is typically between 8 and 12 millimeters. To ensure that the sequence of cross-sectional positions can fully cover the distribution of rib protrusions and depressions, the preset interval is set to 1 / 3 to 1 / 2 of the rib spacing. Given a known image resolution, the millimeter interval is converted to pixel units: if the image resolution is R pixels / millimeter and the rib spacing is P millimeters, then the preset interval is D... pixel =(P / 3)×R to (P / 2)×R pixels.

[0031] For example, when the rib spacing is 10 mm and the image resolution is 2 pixels / mm, the preset interval distance range is 7 to 10 pixels, and in practical applications, 8 pixels can be used as the interval distance.

[0032] Specifically, the selection of the interval distance is related to the periodic undulation of the threaded ribs on the surface of the rebar. The surface of the threaded rebar has regularly distributed transverse ribs, and the spacing between adjacent ribs constitutes an undulation period. If the cross-section position falls precisely on a rib protrusion, the edge width of that cross-section will be greater than the actual diameter of the rebar; if the cross-section position falls on a rib concavity, the edge width will be less than the actual diameter. The preset interval distance is less than the rib periodic spacing, ensuring that the cross-section position sequence includes multiple cross-sections falling on both protrusions and concavities. Subsequent averaging calculations offset the local deviations caused by the threaded undulations. Based on the above cross-section position sequence, the boundary pixel coordinates of the rebar contour are extracted for each cross-section position. The cross-section position corresponds to a scan line perpendicular to the length direction of the rebar in the image. The scan line intersects the rebar contour at two boundary points, namely the left boundary and the right boundary.

[0033] It should be noted that the number of pixels between the left and right boundary pixel coordinates represents the edge width corresponding to that section location. The number of pixels is calculated by subtracting the left boundary pixel coordinate from the right boundary pixel coordinate and taking the absolute value. The resulting value, expressed in pixels, represents the projected width of the reinforcing bar at that section.

[0034] In one embodiment, after summing the edge widths at each cross-section location to form an edge width sequence, the arithmetic mean of all edge widths in the sequence is calculated. The resulting mean eliminates the local fluctuations caused by the protrusions and depressions of the threaded ribs and is used as the actual measured diameter output of the reinforcing bar.

[0035] S104. Based on the outline edge and measured diameter of each steel bar, extract the centerline position of each steel bar and determine the pair combination of adjacent steel bars according to the steel bar arrangement order.

[0036] Based on the outline edge and measured diameter of a single rebar, the left and right boundary pixel coordinates of the outline edge are extracted. The arithmetic mean of the left and right boundary pixel coordinates is taken as the center point coordinates of the cross section. The center point coordinates of each cross section are connected along the length of the rebar to form the centerline position of the rebar. Based on the centerline position of each rebar, the coordinate value of the centerline of each rebar in the arrangement direction is extracted. The rebars are sorted in ascending order of the coordinate values ​​to obtain the rebar arrangement order. Based on the rebar arrangement order, two adjacent rebars are marked as a pair of adjacent rebars. All adjacent rebars are traversed sequentially according to the arrangement order to obtain the pairwise combinations of adjacent rebars. The centerline position and measured diameter of each pair of adjacent rebars are retrieved, and the outer surface outline edge position of the two rebars on the side closest to each other is identified. The net spacing between the surfaces of adjacent rebars is obtained based on the gap width between the outer surface outline edges. This method serves as a supplementary calculation path based on centerline positioning.

[0037] In one implementation, the outline edge of a single rebar is represented in the image as the left and right boundary lines, with the pixel coordinates of the left and right boundaries corresponding to the outer contour positions of the rebar projection. Due to the influence of the image acquisition angle and the threads on the rebar surface, the position of the outline edge may fluctuate locally. Directly taking the midpoint of the left and right boundaries as the center line position would introduce deviations caused by edge fluctuations.

[0038] Specifically, the center point coordinates are calculated by the arithmetic mean of the left boundary pixel coordinates and the right boundary pixel coordinates, using the formula: x center =(x left +x right ) / 2, where x left x represents the pixel coordinates of the left boundary. right The coordinates are the pixel coordinates of the right boundary. After calculating the center point coordinates of each section along the length of the rebar, these center point coordinates are connected in spatial order to form the centerline position of the rebar. Based on the above centerline position, the position coordinate value of the centerline of each rebar in the direction perpendicular to the length of the rebar is extracted. In the scenario where the rebars are arranged in parallel, this position coordinate value reflects the relative position of the rebars in the arrangement direction.

[0039] It should be noted that after sorting the steel bars according to their position coordinate values ​​from smallest to largest, the sorting result is the arrangement order of the steel bars from one side to the other in the image.

[0040] In one embodiment, according to the arrangement order of the reinforcing bars, two reinforcing bars with adjacent sorting positions are marked as a pair of adjacent reinforcing bars. All adjacent positions in the arrangement order are traversed sequentially to obtain a pair of adjacent reinforcing bars. Each pair of pairs contains the centerline position of the two reinforcing bars and their respective measured diameter information.

[0041] S105. Extract the distance between the centerlines of adjacent reinforcing bars as the center spacing based on their centerline positions. At the same time, subtract the radii on both sides based on the measured diameter to obtain the net spacing between the surfaces of adjacent reinforcing bars.

[0042] Based on the centerline positions of each pair of adjacent reinforcing bars, the distance between the centerlines of the two reinforcing bars along the arrangement direction is extracted, and this distance is taken as the center-to-center spacing of the pair of adjacent reinforcing bars. Based on the center-to-center spacing and the measured diameters of the two reinforcing bars, the sum of half the measured diameters of the two reinforcing bars is subtracted from the center-to-center spacing to obtain the theoretical net spacing calculated based on the centerlines. To obtain a more accurate net spacing measurement, the outer surface contour edge positions of the two reinforcing bars on their closest sides are further identified, and the measured net spacing is obtained based on the gap width between the outer surface contour edges.

[0043] In one implementation, the centerline positions of each pair of adjacent reinforcing bars correspond to two parallel or approximately parallel straight lines in the image coordinate system. The distance between the two centerlines along the direction perpendicular to the length of the reinforcing bar is the center-to-center spacing. The center-to-center spacing reflects the geometric interval between the axes of adjacent reinforcing bars and is a direct quantitative indicator of the reinforcing bar density.

[0044] Specifically, the net spacing is calculated based on the relationship between the center-to-center spacing and the measured diameters of the two reinforcing bars. Since the center-to-center spacing measures the distance between the axes of the two reinforcing bars, and the actual gap between the surfaces of the reinforcing bars needs to be reduced by the radius space occupied by each reinforcing bar, the net spacing is equal to the center-to-center spacing minus the sum of half the measured diameters of the two reinforcing bars.

[0045] It should be noted that the net spacing refers to the width of the gap between the outer surfaces of adjacent reinforcing bars. This value is directly related to whether the aggregate can pass smoothly through the gap between the reinforcing bars during concrete pouring, and is a key parameter for construction quality acceptance.

[0046] The centerline position and measured diameter of each pair of adjacent steel bars are retrieved. The spacing between the centerlines of the two steel bars along the arrangement direction is extracted as the center spacing. The outer surface contour edge position of the two steel bars on the side that is close to each other is identified. The net spacing between the surfaces of adjacent steel bars is obtained according to the gap width between the outer surface contour edges.

[0047] The centerline position and measured diameter of each pair of adjacent reinforcing bars are retrieved. Based on the centerline position, the coordinate difference between the centerlines of the two reinforcing bars along the arrangement direction is extracted, and the absolute value of the coordinate difference is taken as the center-to-center distance of the pair of adjacent reinforcing bars. Based on the center-to-center distance and the centerline position of the two reinforcing bars, the outer surface contour edge positions on the sides of the two reinforcing bars that are close to each other are identified. The outer surface contour edge position on the close side of one reinforcing bar is the contour boundary of that reinforcing bar towards the other reinforcing bar, and the outer surface contour edge position on the close side of the other reinforcing bar is the contour boundary of that reinforcing bar towards the previous reinforcing bar. Based on the outer surface contour edge positions on the close side of the two reinforcing bars, the coordinate difference between the two outer surface contour edges in the arrangement direction is extracted, and the absolute value of the coordinate difference is taken as the gap width between the outer surface contour edges. Based on the gap width, the gap width is taken as the net distance between the surfaces of adjacent reinforcing bars. The center-to-center distance and net distance of each pair of adjacent reinforcing bars are summarized to obtain the distance measurement results for all pairs.

[0048] In one implementation, the centerline position and measured diameter of each pair of adjacent reinforcing bars are retrieved as the basic data for spacing measurement. The centerline position is represented in the form of image coordinates, and the measured diameter is recorded in the form of pixel count. The coordinate difference between the centerlines of two reinforcing bars along the arrangement direction is obtained by extracting the coordinate components of their respective centerlines in the arrangement direction and calculating the difference. The absolute value of this coordinate difference is the center-to-center spacing.

[0049] Specifically, the center-to-center spacing reflects the geometric distance between the axes of two reinforcing bars and is a fundamental measure of the density of reinforcing bar arrangement. In parallel reinforcing bar groups, the center-to-center spacing remains relatively stable along the arrangement direction, while in areas with inclination or bending, the center-to-center spacing changes with the cross-sectional position.

[0050] It should be noted that the measured net spacing is obtained by directly identifying the edge position of the outer surface contour. Compared with the theoretical calculation method based on the center spacing minus the radii on both sides, this method has the advantage of reflecting the true contour shape of the rebar surface, especially when there are periodic rib protrusions on the surface of threaded rebars. If half of the measured diameter is used as the radius for theoretical calculation, the resulting net spacing is an estimate based on the average width of the rebar. However, by directly identifying the edge position of the outer surface contour, the measured gap width can reflect the influence of the rib protrusions on the actual gap. When the rib protrusions of two rebars are exactly opposite each other, the measured net spacing will be smaller than the theoretical net spacing. By directly identifying the edge position of the outer surface contour of two rebars that are close to each other, the measured gap width can reflect the influence of the rib protrusions on the actual gap. When the rib protrusions of two rebars are exactly opposite each other, the actual gap will be smaller than the net spacing calculated based on the average diameter. Based on the above principle, the process of identifying the outer surface contour edge position of two reinforcing bars on their closest side is as follows: For a pair of adjacent reinforcing bars, the relative position of the two reinforcing bars is determined according to the coordinate values ​​of their respective centerlines in the arrangement direction. The reinforcing bar with the smaller coordinate value is located on the front side in the arrangement direction, and the reinforcing bar with the larger coordinate value is located on the rear side in the arrangement direction. The contour boundary of the front reinforcing bar towards the rear reinforcing bar is the outer surface contour edge position of the closest side of the reinforcing bar, and the contour boundary of the rear reinforcing bar towards the front reinforcing bar is the outer surface contour edge position of the closest side of the reinforcing bar.

[0051] In one embodiment, the extraction of the edge position of the outer surface contour near the rebar is based on the boundary pixel coordinates within the rebar contour range. Within the contour range of the front rebar, the boundary pixel with the largest coordinate value in the arrangement direction is the edge of the outer surface contour near the rebar; within the contour range of the rear rebar, the boundary pixel with the smallest coordinate value in the arrangement direction is the edge of the outer surface contour near the rebar. Further, the coordinate difference between the two outer surface contour edges in the arrangement direction reflects the actual gap between the surfaces of the two rebars. The absolute value of this coordinate difference is the gap width, which is expressed in pixels as the distance between the outer surfaces of adjacent rebars.

[0052] Understandably, the gap width, as a measurement of the net spacing, is derived directly from the actual position of the outer surface contour edge in the image, rather than a theoretical calculation based on the diameter. In areas where there are local protrusions or deformations on the rebar surface, this measurement method can capture the actual narrowing of the gap.

[0053] In one embodiment, the center-to-center spacing and net spacing of each pair of adjacent reinforcing bars are summarized to form a pair-by-pair spacing measurement result. Each pair includes the identification information of the two reinforcing bars, the center-to-center spacing value, and the net spacing value. This measurement result provides a data basis for subsequent spacing uniformity assessment and compliance determination.

[0054] S106. Extract the net spacing repeatedly along the length of the reinforcing bar at multiple sections, evaluate whether the net spacing of the same pair of adjacent reinforcing bars is consistent at different sections, identify the uneven spacing areas caused by the inclination or bending of the reinforcing bars, and take the average value of the sections of the uneven areas as the corrected net spacing.

[0055] For the same pair of adjacent reinforcing bars, multiple cross-sectional positions are divided along the length of the reinforcing bars. The net spacing value of the pair of adjacent reinforcing bars is extracted at each cross-sectional position, and the net spacing values ​​of all cross-sectional positions are summarized to form a net spacing sequence for the pair of adjacent reinforcing bars. Based on the net spacing sequence, the global mean of all net spacing values ​​in the sequence is calculated as the benchmark value for deviation detection. For each cross-sectional position, the difference between the net spacing value and the global mean is calculated, and the absolute value of the difference is taken as the deviation amplitude of that cross-sectional position. The deviation amplitude is compared with a preset deviation threshold. If the deviation amplitude exceeds the preset deviation threshold, the cross-sectional position is determined to belong to a non-uniform segment. Based on the distribution of the non-uniform segments along the length of the reinforcing bars, continuous cross-sectional ranges are identified, and multiple non-uniform cross-sectional positions continuously distributed along the length direction are marked as the same non-uniform segment. For the non-uniform segment, the net spacing value of each cross-section position in the segment is extracted, and the arithmetic mean of the net spacing value is calculated as the corrected net spacing to connect global detection and achieve local optimization. The corrected net spacing replaces the value of the corresponding segment in the original net spacing sequence to obtain the corrected net spacing measurement result.

[0056] In one embodiment, multiple cross-sectional positions exist along the length of the same pair of adjacent reinforcing bars, each cross-sectional position corresponding to a net spacing value. By dividing the cross-sectional positions along the length of the reinforcing bars according to a preset interval, the net spacing value of each cross-sectional position is extracted, and the net spacing values ​​of all cross-sectional positions are arranged in spatial order to form a net spacing sequence for the pair of adjacent reinforcing bars. The length of the net spacing sequence is the same as the number of cross-sectional positions, and each element in the sequence corresponds to a net spacing measurement value for one cross-sectional position.

[0057] Specifically, the numerical fluctuations of each element in the net spacing sequence reflect the variation characteristics of the gap between adjacent rebar surfaces along the length direction. Under ideal parallel rebar arrangement, the net spacing values ​​at each cross-section should be consistent, and the net spacing sequence exhibits a horizontal distribution. When the rebar is inclined or bent, the net spacing values ​​at different cross-section locations will differ, and the net spacing sequence exhibits a non-horizontal distribution. Based on the above net spacing sequence, the consistency assessment process is as follows: Calculate the arithmetic mean of all net spacing values ​​in the sequence as the benchmark value. For each cross-section location, calculate the difference between the net spacing value and the benchmark value. The absolute value of this difference is the deviation amplitude at that cross-section location. The deviation amplitude reflects the degree of deviation between the net spacing at a single cross-section location and the overall average level; a larger deviation amplitude indicates that the spacing at that cross-section location deviates more from the overall level.

[0058] It should be noted that the preset deviation threshold is a critical value for judging whether the spacing is uniform, and this threshold is determined according to the allowable spacing deviation range in the construction specifications. When the deviation at a certain cross-section exceeds the preset deviation threshold, that cross-section is determined to be a non-uniform spacing location. All elements in the net spacing sequence are traversed, and cross-sections with deviations exceeding the threshold are marked, forming a set of non-uniform spacing locations. Furthermore, the identification of non-uniform sections is based on the spatial distribution characteristics of non-uniform spacing locations along the length of the reinforcing bars. If multiple non-uniform spacing locations are spatially consecutive and adjacent, these locations constitute a non-uniform section; if a non-uniform spacing location is not spatially adjacent to other non-uniform locations, that location constitutes a separate non-uniform section. By analyzing the continuity of non-uniform spacing locations, all non-uniform spacing locations are divided into several non-uniform sections.

[0059] In one embodiment, the non-uniform section contains multiple consecutive cross-sectional locations, the net spacing values ​​of which all deviate from the overall mean. For each non-uniform section, the net spacing values ​​of all cross-sectional locations within that section are extracted, and the arithmetic mean of these net spacing values ​​is calculated as the corrected net spacing. The reason for using the arithmetic mean is that fluctuations in net spacing within a section may be caused by local deformation of the reinforcing steel; taking the mean can smooth out measurement deviations caused by local fluctuations.

[0060] Understandably, after the corrected net spacing replaces the values ​​of the corresponding segments in the original net spacing sequence, the net spacing values ​​of the non-uniform segments in the net spacing sequence become a uniform corrected mean, while the net spacing values ​​of the uniform segments remain unchanged from the original measurements. The corrected net spacing sequence is output as the final measurement result, which retains the original measurement accuracy of the uniform segments while also providing reasonable correction for the non-uniform segments.

[0061] For example, in actual construction sites, inaccurate positioning during rebar tying may cause local rebar tilting, and the net spacing between adjacent rebars in the tilted area will gradually change along the length direction. Through the multi-section evaluation and correction processing of this solution, these tilted areas can be identified and the corrected net spacing measurement results can be output.

[0062] S107. Integrate the net spacing and center spacing of each adjacent steel bar, compare them with the allowable spacing range specified in the construction drawings, mark the areas that exceed the allowable range, generate steel bar spacing compliance analysis results, and complete the automatic analysis of steel bar spacing under low light conditions.

[0063] The net spacing and center-to-center spacing of each pair of adjacent reinforcing bars are integrated. The allowable spacing range specified in the construction drawings is retrieved, and the net spacing and center-to-center spacing are compared with the allowable range. If they exceed the allowable range, the pair of reinforcing bars is marked as an out-of-limit area. The location information and out-of-limit values ​​of the out-of-limit areas are summarized to generate the reinforcing bar spacing compliance analysis results.

[0064] In one implementation, the net spacing and center-to-center spacing of each pair of adjacent reinforcing bars are integrated and summarized as the final data of the spacing measurement, with each pair of adjacent reinforcing bars corresponding to a set of net spacing and center-to-center spacing values. The construction drawings specify the allowable range of reinforcing bar spacing, which includes the upper and lower limits of the net spacing and the upper and lower limits of the center-to-center spacing.

[0065] Specifically, the net spacing of each pair of adjacent reinforcing bars is compared with the upper and lower limits of the allowable net spacing range. If the net spacing is less than the lower limit or greater than the upper limit, the net spacing of that pair of adjacent reinforcing bars exceeds the allowable range. Similarly, the center-to-center spacing is compared with the upper and lower limits of the allowable center-to-center spacing range to determine whether the center-to-center spacing exceeds the allowable range. Any pair of adjacent reinforcing bars whose net spacing or center-to-center spacing exceeds the allowable range is marked as an out-of-limit area.

[0066] In one embodiment, after summarizing the location information and specific over-limit values ​​of all over-limit areas, a rebar spacing compliance analysis result is formed. This result includes the spatial distribution of over-limit areas, the type of over-limit, and the extent of over-limit, thereby enabling accurate assessment and optimization suggestions for rebar spacing under low-light conditions.

[0067] If the technical solution of this application involves personal information, the product using this solution has clearly informed the user of the personal information processing rules and obtained the user's voluntary consent before processing the personal information. If sensitive personal information is involved, the user's separate consent has been obtained before processing, and the "express consent" requirement is met. For example, a clear sign is placed at the collection device such as a camera to inform the user that they have entered the collection area, and the user's voluntary entry is considered as consent; or the processing device clearly indicates the processing rules and obtains authorization through pop-up windows or by asking the user to upload information themselves. The personal information processing rules include the processor, the purpose of processing, the processing method, and the types of personal information.

[0068] Obviously, those skilled in the art can make various modifications and variations to the embodiments of this application without departing from the spirit and scope of the embodiments of this application. Therefore, if these modifications and variations to the embodiments of this application fall within the scope of the claims of this application and their equivalents, this application also intends to include these modifications and variations.

Claims

1. An automatic analysis method for rebar spacing under low light conditions, characterized in that, include: In low-light construction sites, supplementary lighting units and high-sensitivity cameras are deployed to collect supplementary lighting data on areas where steel bars are arranged, thereby obtaining original images containing the complete arrangement of steel bars. The original image is subjected to brightness enhancement and noise reduction processing to identify the boundary between the rebar area and the background area, extract the contour range of each rebar, and obtain the separation result of a single rebar and the number of rebars. Based on the outline range of each steel bar, the edge width of the steel bar is extracted from multiple sections along the length of the steel bar, and the average value of the edge width of each section is taken as the actual measured diameter of the steel bar. Based on the outline edge of each steel bar and the measured diameter, the centerline position of each steel bar is extracted, and the pair combination of adjacent steel bars is determined according to the steel bar arrangement order. The distance between the centerlines of adjacent steel bars is extracted as the center spacing based on the centerline positions of the adjacent steel bars. At the same time, the net spacing between the surfaces of adjacent steel bars is obtained by subtracting the radii on both sides based on the measured diameter. The net spacing is repeatedly extracted at multiple sections along the length of the reinforcing bar to evaluate whether the net spacing of the same pair of adjacent reinforcing bars is consistent at different sections. Uneven spacing areas caused by the inclination or bending of the reinforcing bars are identified, and the average value of the cross sections of the uneven areas is taken as the corrected net spacing. By integrating the net spacing and center-to-center spacing of adjacent reinforcing bars, comparing them with the allowable spacing range specified in the construction drawings, marking areas that exceed the allowable range, and generating a reinforcing bar spacing compliance analysis result.

2. The automatic analysis method for rebar spacing under low light conditions according to claim 1, characterized in that, The deployment of supplementary lighting units and high-sensitivity cameras at the low-light construction site to supplement lighting and collect images of the rebar arrangement area yields original images containing the complete rebar arrangement, including: Based on the ambient light intensity at the construction site, multiple sets of supplementary lighting units are arranged above and to the side of the steel bar arrangement area. The irradiation angle and output brightness of the supplementary lighting units are adjusted so that the irradiation range covers the entire surface of the steel bar arrangement area. A high-sensitivity camera is used to acquire images of the area where the steel bars are arranged under supplementary lighting conditions. The photosensitive parameters of the camera are adapted and adjusted according to the brightness of the supplementary lighting conditions to obtain an original image containing the complete arrangement of the steel bars.

3. The automatic analysis method for rebar spacing under low light conditions according to claim 1, characterized in that, The process of enhancing brightness and denoising the original image, identifying the boundary between the rebar area and the background area, extracting the contour range of each rebar, and obtaining the separation result of a single rebar and the number of rebars includes: Obtain the grayscale distribution of the original image, and perform brightness enhancement processing on the original image based on the proportion of low-brightness pixels in the grayscale distribution to obtain the enhanced image; The enhanced image is denoised using Gaussian filtering to obtain a preprocessed image; Based on the grayscale distribution of the preprocessed image, the Otsu thresholding method is used to determine the boundary threshold between the steel reinforcement area and the background area, resulting in a binary segmentation image. Connected component marking is performed on the rebar region in the binary segmentation image. The contour range of each connected component is extracted as the contour boundary of a single rebar. The number of connected components is counted to obtain the number of rebars. The separation result of a single rebar and the number of rebars are then output.

4. The automatic analysis method for rebar spacing under low light conditions according to claim 1, characterized in that, The step of extracting the edge width of each reinforcing bar from multiple cross-sections along its length based on the outline range of each reinforcing bar, and taking the average of the edge widths of each cross-section as the measured diameter of the reinforcing bar, includes: Based on the outline range of a single steel bar, determine the starting and ending positions of the outline along the length of the steel bar, and divide multiple cross-sectional positions according to a preset interval. For each cross-section location, extract the left boundary pixel coordinates and right boundary pixel coordinates where the cross-section intersects with the steel bar outline, and calculate the number of pixels between the two as the edge width corresponding to the cross-section; Summarize the edge widths at each cross-section location, calculate the arithmetic mean of all edge widths, and use the arithmetic mean as the measured diameter of the reinforcing bar.

5. The automatic analysis method for rebar spacing under low light conditions according to claim 1, characterized in that, The process of extracting the centerline position of each reinforcing bar based on its outline edge and measured diameter, and determining the pairwise combination of adjacent reinforcing bars according to their arrangement order, includes: Based on the outline edge of a single steel bar and the measured diameter, the left boundary pixel coordinates and the right boundary pixel coordinates of the outline edge are extracted, and the arithmetic mean of the two is taken as the center point coordinates of the cross section. The center point coordinates of each cross section are connected to form the center line position of the steel bar. Based on the centerline position of each steel bar, the coordinate value of the centerline of each steel bar in the arrangement direction is extracted, and the steel bars are sorted in ascending order of the coordinate values ​​to obtain the steel bar arrangement order; Based on the arrangement order of the reinforcing bars, two adjacent reinforcing bars are marked as a pair of adjacent reinforcing bars, and the pair combinations of adjacent reinforcing bars are obtained by iterating through them in turn.

6. The automatic analysis method for rebar spacing under low light conditions according to claim 1, characterized in that, The process of extracting the distance between the centerlines of adjacent reinforcing bars as the center-to-center spacing based on their centerline positions, and simultaneously subtracting the radii from both sides based on the measured diameter to obtain the net spacing between the surfaces of adjacent reinforcing bars, includes: Based on the centerline position of each pair of adjacent reinforcing bars, the distance between the centerlines of the two reinforcing bars along the arrangement direction is extracted as the center-to-center spacing of the pair of adjacent reinforcing bars. Based on the center-to-center distance and the measured diameters of the two reinforcing bars, the sum of half the measured diameters of the two reinforcing bars is subtracted from the center-to-center distance to obtain the theoretical net distance calculated based on the centerline. Identify the position of the outer surface contour edge of two steel bars that are close to each other, and obtain the measured net spacing based on the gap width between the outer surface contour edges.

7. The automatic analysis method for rebar spacing under low light conditions according to claim 6, characterized in that, The process of identifying the outer surface contour edge positions of two reinforcing bars on their closest sides, and obtaining the net spacing between adjacent reinforcing bar surfaces based on the gap width between the outer surface contour edges, includes: Based on the centerline position of each pair of adjacent steel bars and their respective measured diameters, the coordinate difference between the centerlines of the two steel bars along the arrangement direction is extracted, and the absolute value of the coordinate difference is taken as the center spacing. Based on the center spacing and centerline position, determine the outer surface contour edge position of the two reinforcing bars on the side that is close to each other. The outer surface contour edge position of one reinforcing bar on the side that is close to each other is the contour boundary of the reinforcing bar towards the other reinforcing bar, and the outer surface contour edge position of the other reinforcing bar on the side that is close to each other is the contour boundary of the reinforcing bar towards the previous reinforcing bar. Extract the coordinate difference between the two outer surface contour edges in the arrangement direction, and use the absolute value of the coordinate difference as the gap width between the outer surface contour edges; The gap width is used as the net spacing between adjacent steel bar surfaces.

8. The automatic analysis method for rebar spacing under low light conditions according to claim 1, characterized in that, The process involves repeatedly extracting the net spacing along the length of the reinforcing bars at multiple cross-sections, evaluating whether the net spacing of the same pair of adjacent reinforcing bars is consistent across different cross-sections, identifying areas of uneven spacing caused by the inclination or bending of the reinforcing bars, and taking the cross-sectional mean value of the uneven areas as the corrected net spacing. This includes: For the same pair of adjacent reinforcing bars, multiple cross-sectional positions are divided along the length of the reinforcing bars. The net spacing value of the pair of adjacent reinforcing bars is extracted at each cross-sectional position to form a net spacing sequence. Based on the net spacing sequence, the global mean of all net spacing values ​​is calculated as the deviation detection benchmark value; For each cross-sectional location, calculate the difference between the net spacing value and the global mean, and take the absolute value of the difference as the deviation amplitude; The deviation amplitude is compared with a preset deviation threshold. If the deviation amplitude exceeds the deviation threshold, the cross-sectional position is determined to be an uneven section. Based on the distribution of non-uniform sections along the length of the reinforcing bars, multiple consecutive non-uniform sections are marked as the same non-uniform section. For the non-uniform segment, the arithmetic mean of the net spacing values ​​of each cross-section position within the segment is calculated as the corrected net spacing, and the value of the corresponding segment in the original net spacing sequence is replaced.

9. The automatic analysis method for rebar spacing under low light conditions according to claim 1, characterized in that, The process integrates the net spacing and center-to-center spacing of adjacent reinforcing bars, compares them with the allowable spacing range specified in the construction drawings, marks areas exceeding the allowable range, and generates a reinforcing bar spacing compliance analysis result, including: Integrate the net spacing and center-to-center spacing of each pair of adjacent reinforcing bars, and retrieve the allowable spacing range specified in the construction drawings; Compare the net spacing and center spacing with the allowable spacing range respectively. If they exceed the allowable spacing range, mark the pair of reinforcing bars as an over-limit area. The location information and excess values ​​of all areas exceeding the limit are summarized to generate the compliance analysis results of the rebar spacing.

Images