Duct piece defect detection method and device based on laser point cloud and medium

Abstract

The embodiment of the invention discloses a segment defect detection method and device based on laser point cloud and a medium, and relates to the technical field of laser point cloud.The method comprises the steps that current scanning point cloud data corresponding to a to-be-detected concrete segment and standard point cloud data corresponding to a standard concrete segment are obtained; determining at least one duct piece target detection position corresponding to the to-be-detected concrete duct piece, and determining a specified standard point corresponding to the duct piece target detection position in the standard point cloud data based on each duct piece target detection position; according to the specified standard point and a preset normal vector matching rule, point cloud data searching is carried out in the current scanning point cloud data, and a current scanning corresponding point corresponding to the specified standard point is determined; and performing defect detection on the target detection position of the segment through the current scanning corresponding point and the specified standard point, and determining defect detection data of the target detection position of each segment. The points matched with the specified standard points are quickly screened out, unnecessary global search is reduced, and the detection efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the field of laser point cloud technology, and in particular to a method, device and medium for detecting pipe segment defects based on laser point cloud. Background Art

[0002] Concrete segments are often used in the construction of new energy and infrastructure such as wind turbine towers and underground pipelines. These facilities have extremely high requirements for structural safety. Defect detection can promptly detect cracks, holes, deformations and other problems in the segments, thereby avoiding potential safety hazards and ensuring the stable operation of the facilities. With the application of 3D laser scanners in reverse engineering, industrial manufacturing, engineering construction, and medical visualization, people have been able to easily obtain a variety of raw point cloud data at a low cost as a basis for further processing. Point cloud data can provide extremely high-precision three-dimensional coordinate information, which can accurately capture the shape, size and surface details of concrete segments. High-precision measurement capabilities are essential for accurately detecting tiny defects in the segments. In addition, the use of point cloud data for detection does not require direct contact with the concrete segments, avoiding possible damage or contamination to the segments.

[0003] At present, defect detection is usually carried out through global point cloud matching. Due to the complex surface structure and large volume of concrete segments, the data of laser scanning points is very large, resulting in uncertainty as to whether a point cloud exists at a certain location; in addition, the amount of point cloud data is huge, and global search consumes a lot of time. Therefore, due to the particularity of the concrete segment structure and the huge amount of point cloud data, the current defect detection method cannot determine the point cloud matching relationship of a certain location point, resulting in inaccurate defect detection results and low matching efficiency. Summary of the invention

[0004] One or more embodiments of the present specification provide a segment defect detection method, device and medium based on laser point cloud, which are used to solve the following technical problems: Due to the particularity of the concrete segment structure, the amount of point cloud data is huge, and the current defect detection method cannot determine the point cloud matching relationship of a certain location point, resulting in inaccurate defect detection results and low matching efficiency.

[0005] One or more embodiments of this specification adopt the following technical solutions:

[0006] One or more embodiments of the present specification provide a segment defect detection method based on laser point cloud, the method comprising: obtaining current scanning point cloud data corresponding to a concrete segment to be detected and standard point cloud data corresponding to a standard concrete segment; determining at least one segment target detection position corresponding to the concrete segment to be detected, and determining, based on each segment target detection position, a designated standard point corresponding to the segment target detection position in the standard point cloud data; performing a point cloud data search in the current scanning point cloud data according to the designated standard point and a preset normal vector matching rule, to determine a current scanning corresponding point corresponding to the designated standard point; performing defect detection on the segment target detection position through the current scanning corresponding point and the designated standard point, and determining defect detection data for each segment target detection position.

[0007] Furthermore, according to the designated standard point and the preset normal vector matching rule, a point cloud data search is performed in the current scanning point cloud data to determine the current scanning corresponding point corresponding to the designated standard point, specifically including: aligning the current scanning point cloud data to the standard point cloud data to determine a search point set in the current scanning point cloud data based on the designated standard point, wherein the search point set includes a plurality of designated scanning point cloud data; obtaining a predetermined reference point, constructing a vector through each of the designated scanning point cloud data and the reference point in the search point set to determine a plurality of matching vectors; determining the current scanning corresponding point corresponding to the designated standard point in the search point set through the plurality of matching vectors and the preset reference vector.

[0008] Furthermore, the current scanning corresponding point corresponding to the designated standard point is determined in the search point set through the multiple matching vectors and the pre-set reference vectors, specifically including: calculating the unit vector dot product according to each of the matching vectors and the reference vector, and determining the multiple vector matching dot products corresponding to the search point set; eliminating point cloud data of the search point set through the multiple vector matching dot products and preset direction judgment rules to determine at least one retained point cloud in the search point set; calculating the distance between each of the retained point clouds and the reference point to determine the corresponding multiple matching distances; and determining the point cloud with the largest matching distance in the at least one retained point cloud through the multiple matching distances to determine the current scanning corresponding point corresponding to the designated standard point.

[0009] Furthermore, point cloud data of the search point set is eliminated through the multiple vector matching dot products and preset direction judgment rules to determine at least one retained point cloud in the search point set, specifically including: obtaining a preset allowable deviation angle to determine a direction judgment threshold through the allowable deviation angle; according to each of the vector matching dot products and the direction judgment threshold, at least one point cloud in the search point set that is smaller than the direction judgment threshold is eliminated to determine at least one retained point cloud in the search point set.

[0010] Furthermore, based on the designated standard point, a search point set is determined in the current scanning point cloud data, specifically including: acquiring a plurality of point cloud data within a designated radius with the designated standard point as the center; determining the data source attributes of each of the plurality of point cloud data, wherein the data source attributes include current scanning point cloud attributes and standard point cloud attributes; based on the data source attributes of each of the point cloud data, screening is performed in the plurality of point cloud data to acquire a plurality of designated scanning point cloud data corresponding to the current scanning point cloud attributes, and determine the search point set.

[0011] Further, obtaining a predetermined reference point specifically includes: determining the reference point according to the designated standard point, wherein a vector between the reference point and the designated standard point and a normal vector of a plane where the designated standard point is located meet preset requirements.

[0012] Furthermore, according to each of the vector matching dot products and the direction judgment threshold, after eliminating at least one point cloud in the search point set that is smaller than the direction judgment threshold, the method also includes: when each of the vector matching dot products in the search point set is smaller than the direction judgment threshold, updating the direction judgment threshold, determining an updated judgment threshold, and eliminating point clouds in the search point set through the updated judgment threshold to determine the at least one retained point cloud.

[0013] Furthermore, determining at least one segment target detection position corresponding to the concrete segment to be detected specifically includes: obtaining a historical defect detection data set corresponding to the concrete segment to be detected, wherein the historical defect detection data set includes historical detection defect parameters and historical detection defect positions; based on the historical detection defect positions in the historical defect detection data set, performing statistical analysis on the defect positions of the concrete segment to determine at least one high-frequency defect position; determining at least one exceeding-standard defect position greater than a preset defect parameter threshold based on multiple historical detection defect parameters in the historical defect detection data set; and determining at least one segment target detection position corresponding to the concrete segment to be detected through the at least one high-frequency defect position and the at least one exceeding-standard defect position.

[0014] One or more embodiments of this specification provide a segment defect detection device based on laser point cloud, including:

[0015] at least one processor; and,

[0016] a memory communicatively connected to the at least one processor; wherein,

[0017] The memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to:

[0018] Acquire current scanning point cloud data corresponding to the concrete segment to be inspected and standard point cloud data corresponding to the standard concrete segment; determine at least one segment target detection position corresponding to the concrete segment to be inspected, so as to determine, in the standard point cloud data, a designated standard point corresponding to the segment target detection position based on each segment target detection position; perform point cloud data search in the current scanning point cloud data according to the designated standard point and a preset normal vector matching rule to determine the current scanning corresponding point corresponding to the designated standard point; perform defect detection on the segment target detection position through the current scanning corresponding point and the designated standard point, and determine the defect detection data of each segment target detection position.

[0019] One or more embodiments of this specification provide a non-volatile computer storage medium storing computer executable instructions, wherein the computer executable instructions are configured to:

[0020] Acquire current scanning point cloud data corresponding to the concrete segment to be inspected and standard point cloud data corresponding to the standard concrete segment; determine at least one segment target detection position corresponding to the concrete segment to be inspected, so as to determine, in the standard point cloud data, a designated standard point corresponding to the segment target detection position based on each segment target detection position; perform point cloud data search in the current scanning point cloud data according to the designated standard point and a preset normal vector matching rule to determine the current scanning corresponding point corresponding to the designated standard point; perform defect detection on the segment target detection position through the current scanning corresponding point and the designated standard point, and determine the defect detection data of each segment target detection position.

[0021] At least one of the above technical solutions adopted in the embodiments of this specification can achieve the following beneficial effects: through the above technical solution, by determining the designated standard point in the standard point cloud data based on the target detection position of the pipe segment, it is possible to ensure that there is an accurate reference point for subsequent point cloud data matching, thereby improving the accuracy of defect detection; using the normal vector matching rule to search for point cloud data not only depends on the spatial position of the point, but also takes into account the local geometric features of the point (i.e., the direction of the surface). The dual matching standard increases the robustness of the search process, so that even in the presence of noise or incomplete alignment of the point cloud data, the corresponding points in the current scan can be found more accurately; through the preset normal vector matching rule, points with similar geometric features to the designated standard points can be quickly screened out, thereby reducing unnecessary global searches, improving detection efficiency, and having Targeted search methods are more efficient than global matching; accurate corresponding point matching and normal vector matching can reduce the possibility of false alarms (misjudging normal points as defect points) and missed alarms (failure to detect actual defect points), ensuring the reliability and accuracy of the detection results; point cloud data matching methods based on rules and algorithms can achieve a high degree of automation, reduce the need for manual intervention, thereby saving time and cost, and improving the consistency and repeatability of the detection process. In addition, normal vector matching rules can be adjusted according to specific detection needs and scenarios to adapt to different types of concrete segments and different degrees of defect detection accuracy requirements; by comparing the current scanned point cloud data with the standard point cloud data, the target detection position on the concrete segment to be detected can be accurately located and compared with the standard segment to accurately detect defects. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of this specification or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art description. Obviously, the drawings described below are only some embodiments recorded in this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. In the drawings:

[0023] Figure 1 A schematic diagram of a flow chart of a segment defect detection method based on laser point cloud provided in an embodiment of this specification;

[0024] Figure 2 A schematic diagram of the structure of a segment defect detection device based on laser point cloud provided in an embodiment of this specification. DETAILED DESCRIPTION

[0025] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this specification.

[0026] Concrete segments are often used in the construction of new energy and infrastructure such as wind turbine towers and underground pipelines. These facilities have extremely high requirements for structural safety. Defect detection can promptly detect cracks, holes, deformations and other problems in the segments, thereby avoiding potential safety hazards and ensuring the stable operation of the facilities. With the application of 3D laser scanners in reverse engineering, industrial manufacturing, engineering construction, and medical visualization, people have been able to easily obtain a variety of raw point cloud data at a low cost as a basis for further processing. Point cloud data can provide extremely high-precision three-dimensional coordinate information, which can accurately capture the shape, size and surface details of concrete segments. High-precision measurement capabilities are essential for accurately detecting tiny defects in the segments. In addition, the use of point cloud data for detection does not require direct contact with the concrete segments, avoiding possible damage or contamination to the segments.

[0027] At present, defect detection is usually carried out through global point cloud matching. Due to the complex surface structure and large volume of concrete segments, the data of laser scanning points is very large, resulting in uncertainty as to whether a point cloud exists at a certain location; in addition, the amount of point cloud data is huge, and global search consumes a lot of time. Therefore, due to the particularity of the concrete segment structure and the huge amount of point cloud data, the current defect detection method cannot determine the point cloud matching relationship of a certain location point, resulting in inaccurate defect detection results and low matching efficiency.

[0028] The embodiment of this specification provides a segment defect detection method based on laser point cloud. It should be noted that the execution subject in the embodiment of this specification can be a server or any device with data processing capabilities. Figure 1 A schematic diagram of a flow chart of a segment defect detection method based on laser point cloud provided in an embodiment of this specification, such as Figure 1 As shown, it mainly includes the following steps:

[0029] Step S101, obtaining current scanning point cloud data corresponding to the concrete segment to be inspected and standard point cloud data corresponding to the standard concrete segment.

[0030] In one embodiment of the present specification, when acquiring point cloud data, the concrete segment can be scanned by a three-dimensional scanning device to acquire the original point cloud data corresponding to the scanned concrete segment. The three-dimensional scanning device here can be a laser scanner or a structured light scanner. In order to ensure the accuracy of the point cloud data, the point cloud data can be preprocessed, for example, the data can be cleaned to remove noise and outliers in the point cloud data to improve the accuracy and reliability of the data. In this process, the current scanned point cloud data corresponding to the concrete segment to be inspected is obtained by scanning and processing the concrete segment to be inspected. In order to perform defect detection on the concrete segment to be inspected, defect-free reference concrete segment data is required. Therefore, the standard concrete segment is scanned and processed to form standard point cloud data of concrete segments of the same specifications and sizes. The standard point cloud data here can be called a standard point cloud model.

[0031] In one embodiment of the present specification, after obtaining the current scanned point cloud data and the standard point cloud data, it is necessary to align the current scanned point cloud data to the standard point cloud data. The currently scanned point cloud data and the standard point cloud data are preprocessed, including removing noise, filling holes, smoothing, etc., to improve the accuracy of the registration, ensure that the two point cloud data sets are in the same coordinate system or perform appropriate coordinate transformation. Afterwards, significant feature points such as corner points, edge points or surface features are extracted from the two point cloud data sets, and the above feature points will be used in the subsequent registration process. Point cloud registration can be divided into two stages: coarse registration and fine registration. If the initial positions of the two point clouds are very different, a rough initial alignment can be performed first. The coarse registration roughly aligns the same-name areas of the point clouds of the two models, i.e., the corresponding surfaces. It can also be achieved by manual adjustment, using a marked object (such as a calibration plate) or an automatic alignment method based on global features. The fine matching criterion uses the iterative closest point (ICP) algorithm or its variants for precise registration. The ICP algorithm iteratively searches for the closest point correspondence between two point clouds and calculates a transformation matrix (including rotation and translation) to minimize the distance between the two point clouds. Through multiple iterations, the algorithm gradually adjusts the position and posture of the current scanned point cloud until the preset convergence condition or the maximum number of iterations is reached. Calculate the error indicators between the two registered point clouds, such as the root mean square error (RMSE) or the Hausdorff distance, to evaluate the quality of the registration. If the error is large, it may be necessary to adjust the registration parameters or try other registration methods. When the registration result meets the preset requirements, the calculated transformation matrix is ​​applied to the currently scanned point cloud data to align it with the standard point cloud.

[0032] Through the above technical solution, the volume of concrete segments is relatively large. When scanning concrete segments, each scan may only capture a part of the concrete segments. Through registration, these partial scanning data can be integrated into a unified coordinate system to form a complete and continuous three-dimensional model. Since there may be various errors in the scanning process (such as equipment errors, environmental interference, etc.), these errors can be corrected through registration to improve the accuracy of the three-dimensional model. The algorithm used in the registration process will calculate and apply the necessary rotation, translation and scaling transformations to minimize the differences between different scanning data. In addition, registration can fuse point cloud data from different perspectives to generate a more comprehensive and detailed three-dimensional representation.

[0033] Step S102, determining at least one segment target detection position corresponding to the concrete segment to be detected, so as to determine, based on each segment target detection position, a designated standard point corresponding to the segment target detection position in the standard point cloud data.

[0034] In conventional defect detection methods, the object to be detected is used as the entry point, and the detection data corresponding to the object to be detected is compared with the standard data one by one to determine the detection result. In the application scenario of defect detection of concrete segments, due to the large volume of concrete segments, if the concrete segments are used as the entry point, the measured detection data is compared with the standard data one by one, which is a lot of work and leads to low detection efficiency; in addition, due to the large amount of point cloud data obtained by laser scanning, it is impossible to determine whether there is scanned point cloud data at a certain position.

[0035] Determining at least one target detection position of a segment corresponding to the concrete segment to be detected specifically includes: obtaining a historical defect detection data set corresponding to the concrete segment to be detected, wherein the historical defect detection data set includes historical detection defect parameters and historical detection defect positions; based on the historical detection defect positions in the historical defect detection data set, performing statistical analysis on the defect positions of the concrete segment to determine at least one high-frequency defect position; determining at least one excessive defect position greater than a preset defect parameter threshold value based on multiple historical detection defect parameters in the historical defect detection data set; and determining at least one target detection position of a segment corresponding to the concrete segment to be detected through the at least one high-frequency defect position and the at least one excessive defect position.

[0036] In one embodiment of the present specification, at least one segment target detection position corresponding to the concrete segment to be detected is determined based on the historical detection results of the concrete segment to be detected. First, multiple concrete segments that have been detected with the same specifications and parameters as the concrete segment to be detected are determined. Secondly, a historical defect detection data set of multiple concrete segments that have been detected is obtained in the detection data storage database, where the historical defect detection data set includes historical defect detection information of multiple concrete segments, specifically including historical detection defect parameters and historical detection defect positions.

[0037] When determining the target detection position of the segment, based on the historical detection defect positions in the historical defect detection data set, the defect positions of the concrete segment are statistically analyzed to determine at least one high-frequency defect position, where high frequency means that the number of defects detected at the same position meets the preset number threshold, and the preset number threshold can be set according to actual needs. Based on multiple historical detection defect parameters in the historical defect detection data set, at least one excessive defect position greater than the preset defect parameter threshold is determined, where the excessive defect position refers to a position with a larger defect, and the defect parameter threshold can be set according to the defect type and actual needs. At least one segment target detection position corresponding to the concrete segment to be detected is determined through at least one high-frequency defect position and at least one excessive defect position, that is, the segment target detection position is a position where more defects occurred during the historical detection process and a position where larger defects occurred.

[0038] When determining the target detection position of the segment, it is also possible to statistically analyze the historical defect detection information of multiple concrete segments, and to count multiple specified defect detection results with defects, so as to determine the segment defect positions corresponding to the multiple specified defect detection results according to the multiple specified defect detection results. The multiple segment positions obtained are screened according to the number of defects, and high-frequency segment defect positions that are higher than the preset defect number threshold are obtained. It should be noted that the preset defect number threshold here can be determined based on the data volume of the historical defect detection data set, for example, the defect number threshold is inversely proportional to the data volume, and can also be set according to the actual defect detection requirements. The obtained high-frequency segment defect position is used to determine the segment target detection position corresponding to the concrete segment to be detected, and is sorted according to the defect number corresponding to the high-frequency segment defect position, so as to facilitate defect detection of each segment target detection position in sequence.

[0039] Through the above technical scheme, by analyzing the historical inspection data of concrete segments of the same specifications, high-frequency defect positions and / or excessive defect positions are screened in the concrete segments as segment target inspection positions. The target positions can be determined by selecting the corresponding method according to actual needs, which effectively reduces the amount of data required for inspection and ensures the representativeness of the less selected target inspection positions, further improving the defect detection efficiency.

[0040] In one embodiment of the present specification, after determining the target detection position of the segment, the designated standard point corresponding to the target detection position of the segment is determined in the standard point cloud data corresponding to the concrete segment according to each target detection position of the segment. In order to determine whether there is actually scanned point cloud data for the target detection position of the segment through the designated standard point. The detection based on the actual scanned point cloud data in the traditional detection method is converted into the detection position point cloud in the standard point cloud data, that is, the designated standard point, which is convenient for finding the corresponding point of the designated standard point in the standard point cloud data in the scanned point cloud in the subsequent process, providing the defect detection process with the result of accurate point cloud matching, and further ensuring the accuracy of defect detection.

[0041] Step S103 , searching the current scan point cloud data according to the designated standard point and the preset normal vector matching rule, so as to determine the current scan corresponding point corresponding to the designated standard point.

[0042] According to the designated standard point and the preset normal vector matching rule, a point cloud data search is performed in the current scanning point cloud data to determine the current scanning corresponding point corresponding to the designated standard point, specifically including: aligning the current scanning point cloud data to the standard point cloud data to determine a search point set in the current scanning point cloud data based on the designated standard point, wherein the search point set includes a plurality of designated scanning point cloud data; obtaining a predetermined reference point, constructing a vector through each of the designated scanning point cloud data and the reference point in the search point set to determine a plurality of matching vectors; determining the current scanning corresponding point corresponding to the designated standard point in the search point set through the plurality of matching vectors and the preset reference vector.

[0043] In one embodiment of the present specification, the current scan point cloud data is registered to the standard point cloud data to obtain a registered point cloud including the scan point cloud and the standard point cloud. The point cloud data is analyzed in the registered point cloud so as to determine the search point set in the current scan point cloud data according to the specified standard point. It should be noted that the search point set includes multiple specified scan point clouds, and each specified scan point cloud belongs to the current scan point cloud data. By searching the point set, the search range of the corresponding point of the specified standard point is narrowed. A predetermined reference point is obtained. It should be noted that the reference point here is located at the other end of the standard point cloud, that is, the reference point belongs to neither the standard point cloud nor the scan point cloud, and is far enough from the standard point cloud, and the vector from the reference point to the specified standard point is as perpendicular as possible to the plane where the specified standard point is located, that is, the reference point is as close as possible to the other end of the normal vector of the plane where the specified standard point is located. A vector is constructed based on each specified scan point cloud data and reference point in the search point set to determine multiple matching vectors. Through multiple matching vectors and pre-set reference vectors, the current scanning corresponding point corresponding to the specified standard point is determined in the search point set. It should be noted that the reference vector here refers to the vector corresponding to the reference point and the point to be searched.

[0044] Based on the designated standard point, a search point set is determined in the scanned point cloud data, specifically including: obtaining a plurality of point cloud data within a designated radius with the designated standard point as the center; determining the data source attribute of each of the plurality of point cloud data in the plurality of point cloud data, wherein the data source attribute includes the current scanned point cloud attribute and the standard point cloud attribute; based on the data source attribute of each of the point cloud data, screening is performed in the plurality of point cloud data to obtain a plurality of designated scanned point cloud data corresponding to the current scanned point cloud attribute, and determine the search point set.

[0045] In one embodiment of the present specification, a designated standard point is taken as the center, and multiple point cloud data within a designated radius range within the designated standard point coordinate range are obtained. The multiple point cloud data include both point clouds in the standard point cloud and point clouds in the scanned point cloud. It should be noted that the designated radius range here is used for clustering, and the value of the radius can be considered from the following aspects. First, at least 3 points can be found within this radius range of all normal points. In addition, there cannot be too many points within this radius range, otherwise it is difficult to distinguish different point clusters. It can be determined based on actual needs and empirical data. Determine the data source attribute of each point cloud data in the multiple point cloud data. The data source attribute here is used to identify the point cloud to which each point belongs. The data source attribute includes the current scan point cloud attribute and the standard point cloud attribute. When the data source attribute of a point cloud in the multiple point cloud data is the current scan point cloud attribute, it means that this point cloud is a point in the current scan point cloud. According to the data source attribute of each point cloud data, multiple point cloud data are screened to obtain multiple specified scan point cloud data corresponding to the current scan point cloud attribute, and the search point set is determined. In other words, the point cloud data in the search point set are all points in the current scan point cloud data.

[0046] Obtaining a predetermined reference point specifically includes: determining the reference point according to a specified standard point, wherein a vector between the reference point and the specified standard point and a normal vector of a plane where the specified standard point is located meet preset requirements.

[0047] In one embodiment of the present specification, the designated standard point can also be understood as a point to be found in the current scanned point cloud data. The defect type corresponding to the segment defect detection is obtained, and based on the defect type, the point to be found is determined in the current scanned point cloud data. The point to be found is the point that the user wants to identify and use on the scanned point cloud. For example, when the defect type is the segment straight edge length detection in the defect detection process, that is, if you want to calculate the length of a straight edge, you need to determine the two endpoints of the edge. Therefore, these two endpoints are determined as designated standard points, which can also be called points to be found. It should be noted that the two endpoints here are called designated standard points in the standard point cloud. In the current scanned point cloud, since it is uncertain whether this point is actually scanned, it can be called a point to be found. According to the designated standard point, the reference point is determined. The vector between the reference point and the point to be found is as perpendicular as possible to the normal vector of the plane where the designated standard point is located. In addition, the reference point is located at the other end of the standard point cloud, that is, as far away as possible from the point to be found, and does not belong to the standard point cloud or the scanned point cloud.

[0048] The current scanning corresponding point corresponding to the designated standard point is determined in the search point set by using the multiple matching vectors and the preset reference vectors, specifically including: calculating the unit vector dot product according to each of the matching vectors and the reference vector, and determining the multiple vector matching dot products corresponding to the search point set; eliminating point cloud data of the search point set by using the multiple vector matching dot products and preset direction judgment rules to determine at least one retained point cloud in the search point set; calculating the distance between each of the retained point clouds and the reference point to determine the corresponding multiple matching distances; and determining the point cloud with the largest matching distance in the at least one retained point cloud by using the multiple matching distances to determine the current scanning corresponding point corresponding to the designated standard point.

[0049] In one embodiment of the present specification, the dot product of the unit vector corresponding to each matching vector and the reference vector is calculated to obtain the vector matching dot products corresponding to the multiple matching vectors of the search point set. Through multiple vector matching dot products and preset direction judgment rules, the point cloud data of the search point set is eliminated to determine at least one retained point cloud in the search point set. It should be noted that when the directions of two vectors are consistent, the dot product of the corresponding unit vectors is 1. Through the direction judgment rule, the point cloud with a large direction deviation is eliminated in the search point set according to the vector matching point set to obtain at least one retained point cloud. After screening the point cloud in the search point set, the distance between each retained point cloud and the reference point is calculated to determine the matching distance corresponding to each retained point cloud. According to the matching distance corresponding to each retained point cloud, the point cloud corresponding to the maximum distance is determined, and this point cloud is determined as the current scanning corresponding point corresponding to the specified standard point. In other words, the outermost points are taken, and it is considered that the outermost points are the points that best conform to the actual situation.

[0050] Through the dot products of the multiple vector matches and the preset direction judgment rules, the point cloud data of the search point set is eliminated to determine at least one retained point cloud in the search point set, specifically including: obtaining a preset allowable deviation angle to determine a direction judgment threshold through the allowable deviation angle; according to each of the vector matching dot products and the direction judgment threshold, at least one point cloud in the search point set that is smaller than the direction judgment threshold is eliminated to determine at least one retained point cloud in the search point set.

[0051] In one embodiment of the present specification, when performing point cloud data elimination, it is necessary to set elimination criteria, and the allowable deviation angle can be set according to requirements or empirical data, that is, two vectors within the allowable deviation angle are retained. First, according to the allowable deviation angle, the direction judgment threshold is determined, and the absolute value of the maximum difference between the allowable unit vector dot product and 1 is set to dev, that is, the direction judgment threshold here can be 1-dev. Assuming that the vector matching dot product is cos, according to the vector matching dot product cos and the direction judgment threshold, the points with cos greater than or equal to 1-dev are retained in the search point set, and at least one point cloud less than 1-dev is eliminated to determine at least one retained point cloud in the search point set.

[0052] After eliminating at least one point cloud smaller than the direction judgment threshold in the search point set according to each vector matching dot product and the direction judgment threshold, the method further includes: when each vector matching dot product in the search point set is smaller than the direction judgment threshold, updating the direction judgment threshold and determining an updated judgment threshold, so as to eliminate point clouds in the search point set through the updated judgment threshold and determine the at least one retained point cloud.

[0053] In one embodiment of the present specification, if the vector matching dot products of all point clouds in the search point set are less than the direction judgment threshold, in order to determine the corresponding point, the value of dev can be doubled to obtain the latest direction judgment threshold, that is, the updated judgment threshold. According to the updated judgment threshold, a new judgment is performed in the search point set in the above manner, and point clouds less than 1-dev are eliminated in the search point set to determine at least one retained point cloud.

[0054] Through the above technical scheme, by registering the scanned point cloud data and the standard point cloud data, the deviation caused by factors such as the scanning angle and the device position can be eliminated, thereby improving the accuracy of locating the specified standard point in the scanned point cloud data; after determining the search point set, by constructing a vector and comparing it with the reference vector, the corresponding point of the specified standard point in the scanned point cloud data can be quickly determined, thereby simplifying the search process and improving efficiency; through registration and vector matching, it can be ensured that the scanned point cloud data and the standard point cloud data are compared in the same coordinate system, which makes the data scanned at different time points or by different devices comparable, which is convenient for subsequent data analysis and processing.

[0055] Step S104, performing defect detection on the target detection position of the segment through the current scanning corresponding point and the designated standard point, and determining the defect detection data of each target detection position of the segment.

[0056] In one embodiment of the present specification, a reasonable deviation threshold is set according to the design requirements and manufacturing process of the concrete pipe segment to determine whether there are defects. Multiple groups of corresponding points are determined, each group of corresponding points includes the current scanned corresponding points and the specified standard points, and the deviation data between the current scanned corresponding points and the corresponding specified standard points in each group of corresponding points is calculated. The deviation here can be a position deviation, a distance deviation, or an angle deviation, which can be determined according to the actual defect detection needs. According to the calculated deviation data, defect detection is performed on each pipe segment target detection position to determine the defect detection data. For example, the calculated deviation data can be compared with the deviation threshold. If the deviation exceeds the threshold, it is determined that there is a defect at the position, and the defect is classified according to the size and type of the deviation (such as bulges, depressions, cracks, etc.). The defects can also be quantified, such as calculating parameters such as the size, depth or length of the defects.

[0057] Through the above technical solution, by determining the designated standard point in the standard point cloud data based on the target detection position of the pipe segment, it is possible to ensure that there is an accurate reference point for subsequent point cloud data matching, thereby improving the accuracy of defect detection; using the normal vector matching rule to search for point cloud data not only depends on the spatial position of the point, but also considers the local geometric features of the point (i.e., the direction of the surface). The dual matching standard increases the robustness of the search process, so that even in the presence of noise or incomplete alignment of the point cloud data, the corresponding points in the current scan can be found more accurately; through the preset normal vector matching rule, points with similar geometric features to the designated standard points can be quickly screened out, thereby reducing unnecessary global searches and improving detection efficiency. Targeted search methods are more accurate than global comparisons. High efficiency; accurate corresponding point matching and normal vector matching can reduce the possibility of false alarms (misjudging normal points as defect points) and missed alarms (failure to detect actual defect points), ensuring the reliability and accuracy of the detection results; the point cloud data matching method based on rules and algorithms can achieve high automation, reduce the need for manual intervention, thereby saving time and cost, and improving the consistency and repeatability of the detection process. In addition, the normal vector matching rules can be adjusted according to specific detection needs and scenarios to adapt to different types of concrete segments and different degrees of defect detection accuracy requirements; by comparing the current scanned point cloud data with the standard point cloud data, the target detection position on the concrete segment to be detected can be accurately located and compared with the standard segment to accurately detect defects.

[0058] The embodiment of this specification also provides a segment defect detection device based on laser point cloud, such as Figure 2 As shown, the device includes: at least one processor; and a memory connected to the at least one processor in communication; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to:

[0059] Acquire current scanning point cloud data corresponding to the concrete segment to be inspected and standard point cloud data corresponding to the standard concrete segment; determine at least one segment target detection position corresponding to the concrete segment to be inspected, and determine the designated standard point corresponding to the segment target detection position in the standard point cloud data based on each segment target detection position; perform point cloud data search in the current scanning point cloud data according to the designated standard point and a preset normal vector matching rule to determine the current scanning corresponding point corresponding to the designated standard point; perform defect detection on the segment target detection position through the current scanning corresponding point and the designated standard point, and determine the defect detection data of each segment target detection position.

[0060] The embodiment of the present specification also provides a non-volatile computer storage medium storing computer executable instructions, wherein the computer executable instructions are configured as follows:

[0061] Acquire current scanning point cloud data corresponding to the concrete segment to be inspected and standard point cloud data corresponding to the standard concrete segment; determine at least one segment target detection position corresponding to the concrete segment to be inspected, and determine the designated standard point corresponding to the segment target detection position in the standard point cloud data based on each segment target detection position; perform point cloud data search in the current scanning point cloud data according to the designated standard point and a preset normal vector matching rule to determine the current scanning corresponding point corresponding to the designated standard point; perform defect detection on the segment target detection position through the current scanning corresponding point and the designated standard point, and determine the defect detection data of each segment target detection position.

[0062] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device, equipment, and non-volatile computer storage medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0063] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0064] The devices and media provided in the embodiments of this specification correspond one-to-one to the methods. Therefore, the devices and media also have similar beneficial technical effects as the corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the devices and media will not be repeated here.

[0065] Those skilled in the art will appreciate that the embodiments of this specification may be provided as methods, systems, or computer program products. Therefore, this specification may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, this specification may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0066] This specification is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of this specification. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0067] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0068] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0069] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0070] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0071] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0072] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0073] The above description is only one or more embodiments of this specification and is not intended to limit this specification. For those skilled in the art, one or more embodiments of this specification may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of one or more embodiments of this specification shall be included in the scope of the claims of this specification.

Claims

1. A segment defect detection method based on laser point cloud, characterized in that: The method comprises: Obtaining current scanning point cloud data corresponding to the concrete segment to be inspected and standard point cloud data corresponding to the standard concrete segment; Determine at least one segment target detection position corresponding to the concrete segment to be detected, and determine, based on each segment target detection position, a designated standard point corresponding to the segment target detection position in the standard point cloud data; According to the designated standard point and the preset normal vector matching rule, searching the point cloud data in the current scan point cloud data to determine the current scan corresponding point corresponding to the designated standard point; Through the current scanning corresponding point and the designated standard point, defect detection is performed on the pipe segment target detection position to determine the defect detection data of each pipe segment target detection position.

2. The method for detecting segment defects based on laser point cloud according to claim 1, characterized in that: According to the designated standard point and the preset normal vector matching rule, searching the point cloud data in the current scan point cloud data to determine the current scan corresponding point corresponding to the designated standard point specifically includes: Aligning the current scan point cloud data to the standard point cloud data to determine a search point set in the current scan point cloud data based on the designated standard point, wherein the search point set includes a plurality of designated scan point cloud data; Obtain a predetermined reference point, and construct a vector through each of the designated scan point cloud data in the search point set and the reference point to determine a plurality of matching vectors; The current scanning corresponding point corresponding to the designated standard point is determined in the search point set by using the multiple matching vectors and the preset reference vector.

3. The method for detecting segment defects based on laser point cloud according to claim 2, characterized in that: Determining the current scanning corresponding point corresponding to the designated standard point in the search point set by using the multiple matching vectors and the preset reference vector specifically includes: Calculate the unit vector dot product according to each of the matching vectors and the reference vector, and determine a plurality of vector matching dot products corresponding to the search point set; By matching dot products of the multiple vectors and a preset direction judgment rule, point cloud data removal is performed on the search point set to determine at least one retained point cloud in the search point set; Calculating the distance between each of the retained point clouds and the reference point to determine a corresponding plurality of matching distances; A point cloud with the largest matching distance is determined in the at least one retained point cloud by using the multiple matching distances to determine the current scanning corresponding point corresponding to the designated standard point.

4. The method for detecting segment defects based on laser point cloud according to claim 3, characterized in that: By matching the dot products of the multiple vectors and a preset direction judgment rule, point cloud data is eliminated from the search point set to determine at least one retained point cloud in the search point set, specifically including: Obtaining a preset allowable deviation angle to determine a direction judgment threshold through the allowable deviation angle; According to each of the vector matching dot products and the direction judgment threshold, at least one point cloud smaller than the direction judgment threshold is eliminated from the search point set to determine at least one retained point cloud in the search point set.

5. The method for detecting segment defects based on laser point cloud according to claim 2, characterized in that: Based on the specified standard point, determining a search point set in the current scan point cloud data specifically includes: Taking the designated standard point as the center, obtaining a plurality of point cloud data within a designated radius; Determine the data source attribute of each of the plurality of point cloud data, wherein the data source attribute includes current scan point cloud attributes and standard point cloud attributes; Based on the data source attribute of each point cloud data, the plurality of point cloud data are screened to obtain a plurality of designated scanning point cloud data corresponding to the current scanning point cloud attribute, and the search point set is determined.

6. The method for detecting segment defects based on laser point cloud according to claim 2, characterized in that: Obtain predetermined reference points, including: The reference point is determined according to the designated standard point, wherein a vector between the reference point and the designated standard point and a normal vector of a plane where the designated standard point is located meet preset requirements.

7. The method for detecting segment defects based on laser point cloud according to claim 4, characterized in that: According to each of the vector matching dot products and the direction determination threshold, after removing at least one point cloud smaller than the direction determination threshold in the search point set, the method further includes: When each vector matching dot product in the search point set is less than the direction judgment threshold, the direction judgment threshold is updated to determine an updated judgment threshold, so as to eliminate point clouds in the search point set through the updated judgment threshold and determine at least one retained point cloud.

8. The method for detecting segment defects based on laser point cloud according to claim 1, characterized in that: Determining at least one segment target detection position corresponding to the concrete segment to be detected specifically includes: Acquire a historical defect detection data set corresponding to the concrete segment to be detected, wherein the historical defect detection data set includes historical detection defect parameters and historical detection defect positions; Based on the historically detected defect positions in the historical defect detection data set, statistically analyzing the defect positions of the concrete segment to determine at least one high-frequency defect position; Determine at least one defect position exceeding a predetermined defect parameter threshold value according to a plurality of historical defect detection parameters in the historical defect detection data set; At least one segment target detection position corresponding to the concrete segment to be detected is determined through the at least one high-frequency defect position and the at least one exceeding-standard defect position.

9. A segment defect detection device based on laser point cloud, characterized in that: The device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to: Obtaining current scanning point cloud data corresponding to the concrete segment to be inspected and standard point cloud data corresponding to the standard concrete segment; Determine at least one segment target detection position corresponding to the concrete segment to be detected, and determine, based on each segment target detection position, a designated standard point corresponding to the segment target detection position in the standard point cloud data; According to the designated standard point and the preset normal vector matching rule, searching the point cloud data in the current scan point cloud data to determine the current scan corresponding point corresponding to the designated standard point; Through the current scanning corresponding point and the designated standard point, defect detection is performed on the pipe segment target detection position to determine the defect detection data of each pipe segment target detection position.

10. A non-volatile computer storage medium storing computer executable instructions, characterized in that: The computer executable instructions are configured to: Obtaining current scanning point cloud data corresponding to the concrete segment to be inspected and standard point cloud data corresponding to the standard concrete segment; Determine at least one segment target detection position corresponding to the concrete segment to be detected, and determine, based on each segment target detection position, a designated standard point corresponding to the segment target detection position in the standard point cloud data; According to the designated standard point and the preset normal vector matching rule, searching the point cloud data in the current scan point cloud data to determine the current scan corresponding point corresponding to the designated standard point; Through the current scanning corresponding point and the designated standard point, defect detection is performed on the pipe segment target detection position to determine the defect detection data of each pipe segment target detection position.

Images