Building engineering quality detection and evaluation system based on machine vision

Abstract

The invention relates to the technical field of image defect detection, in particular to a building engineering quality detection and evaluation system based on machine vision, which comprises an image acquisition module, a crack main shaft identification module, a space packaging module, a structure overlapping detection module and a building quality evaluation module. According to the method, a crack boundary layer with a spatial positioning function is constructed through image edge scanning and pixel position marking, pixel segments with high continuity are screened to generate a path line, the continuity and representativeness of crack trend are improved, and a rectangular block is packaged to enhance organization and complete expression of cracks in a spatial level; performing coordinate comparison with a structure stress area to form corresponding association between defect influence and a structure function, judging the influence of the crack on a key area through a space coverage relation between a total block amount and an overlapping amount, and establishing multi-dimensional building quality grade division by combining crack form, position distribution and structure association information; and comprehensive evaluation of building structure defects from geometry to structure level is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of image defect detection, and in particular to a building engineering quality inspection and evaluation system based on machine vision. Background Art

[0002] The technical field of image defect detection involves using computer vision and image processing methods to automatically identify and analyze defects existing on the surface of an object or inside a structure. The core content is to obtain image data and perform processing means such as feature extraction, image segmentation, edge detection, and pattern recognition on it, to accurately identify and locate defects such as cracks, deformations, and holes in the target object. It is widely used in many fields such as industrial manufacturing, material analysis, safety detection, traffic monitoring, and building structure diagnosis. The overall technological development depends on high-quality image acquisition means, robust image processing algorithms, and image understanding models, aiming to achieve non-contact, high-efficiency, and automated detection of defects.

[0003] Among them, a building engineering quality inspection and evaluation system based on machine vision refers to capturing images of the surface of a building engineering structure by deploying image acquisition devices, and using image analysis methods to identify, extract, and classify defects such as concrete cracks, surface peeling, and honeycombing on the images. It covers image acquisition, lighting control, image preprocessing, defect area extraction, and defect type discrimination. It uses methods such as image edge feature enhancement, crack morphology feature extraction, and regional gray distribution analysis to complete the identification and annotation of building surface defects, and completes the grading of quality evaluation based on the image geometric features of the defect area.

[0004] In the existing building engineering quality inspection and evaluation process, relying on the extraction of image features and the regional division method of static image content, the recognition process mainly focuses on the apparent information in the image, lacking the ability to jointly express the image and the structural spatial distribution, resulting in the recognition result only staying at the primary level of the defect appearance. The defect recognition does not consider the direction trend and path connectivity, resulting in problems such as recognition fracture and morphological fragmentation when dealing with structural extensional defects, and it is difficult to obtain the true evolution trend of cracks in space. The lack of a spatial organization method for encapsulated blocks makes it difficult for the recognition result to achieve coordinate-level mapping with the key areas of the building structure, and it cannot provide effective data support when it is necessary to determine the defect risk in the core area where stress is concentrated. The quality evaluation is based on a single index such as image gray level or area, ignoring the correlation between defects and structural layout, and is prone to evaluation deviation in a complex spatial component environment. For example, in a high-stress concentration area, even if continuous cracks appear, their high-risk attributes cannot be recognized due to the lack of a structural coincidence judgment mechanism, thus affecting the rigor and practicality of the overall quality evaluation. Summary of the Invention

[0005] The object of the present invention is to solve the disadvantages existing in the prior art, and a building engineering quality inspection and evaluation system based on machine vision is proposed.

[0006] To achieve the above object, the present invention adopts the following technical solutions: The building engineering quality inspection and evaluation system based on machine vision includes:

[0007] The image acquisition module acquires the building surface structure image, performs edge scanning on the non-background area of the image, extracts the boundary pixels of the crack area, and marks the position relationship of each pixel in the building image coordinate system to obtain the crack boundary coordinate layer;

[0008] The crack main axis recognition module calculates the edge direction angle value of the point according to the gray change direction of the adjacent pixels of each crack boundary pixel point in the crack boundary coordinate layer, sorts the pixel point sequence and classifies the change trend of the direction angle, constructs the main axis walking path line, and obtains the crack main axis direction line set;

[0009] The space encapsulation module divides the encapsulation rectangular area aligned with the crack direction based on the spatial position relationship of the path lines in the crack main axis direction line set on the building image coordinate, extracts the boundary coordinates of each area, and records the positioning block number of each area to obtain the crack path encapsulation block set;

[0010] The structure overlap detection module compares the spatial boundary coordinates of each block in the crack path encapsulation block set with the boundary data of the stress concentration area in the building structure image to check whether there is a spatial section with overlapping coordinates, summarizes the overlapping block numbers, and obtains the structure block overlap record table.

[0011] As a further solution of the present invention, the crack boundary coordinate layer includes a boundary pixel distribution map, a pixel coordinate relationship table, and a crack image position index. The crack main axis direction line set includes a main axis path line sequence, a direction angle sequence set, and a direction line classification label. The crack path encapsulation block set includes encapsulation rectangular boundary data, path block positioning numbers, and block spatial direction information. The structure block overlap record table includes an overlapping block number table, a coincident spatial section coordinate set, and a stress area correspondence mapping result.

[0012] As a further solution of the present invention, the image acquisition module includes:

[0013] The image structure extraction sub-module acquires the building surface structure image, performs gray conversion processing on the original image data, calculates and distinguishes the gray difference between the background area and the non-background area, selects the background gray threshold as the basis, classifies the area below the background gray threshold as the non-background area, and obtains the non-background pixel distribution map;

[0014] The edge pixel calibrator sub-module locates the edge contours of the non-background area in the image based on the non-background pixel distribution map, judges the gradient direction and gray-scale change amplitude of adjacent pixel points in the image, marks the crack boundaries for pixels with a gray-scale change amplitude greater than the crack recognition gradient threshold, identifies all pixel coordinate information within the boundary contour line, and obtains the crack edge pixel marking result;

[0015] The boundary coordinate generation sub-module converts the positions of all pixel points within the crack boundary relative to the building image coordinate system in the image matrix according to the crack edge pixel marking result, establishes the mapping relationship between the horizontal and vertical coordinate systems of the boundary points, and obtains the crack boundary coordinate layer.

[0016] As a further solution of the present invention, the crack main axis recognition module includes:

[0017] The direction angle extraction sub-module obtains the difference between the gray-scale values of adjacent pixels and the target pixel based on each crack boundary pixel point in the crack boundary coordinate layer, calculates the distribution mean and change rate of the gray-scale difference in the adjacent pixel direction, judges the gray-scale change direction, measures the gradient angle and converts the direction vector for each crack boundary pixel point, and obtains the edge direction angle value;

[0018] The sequence trend classification sub-module analyzes the connectivity relationship of the crack boundary pixel points according to the edge direction angle value, sorts the connected pixel point sequences in adjacent order, calculates the direction angle change rate of each pixel point and the difference in direction angle from the adjacent point, classifies the direction angle change trend, and uses the formula:

[0019] ;

[0020] Performs operations to obtain the direction fluctuation trend value , filters out continuous pixel segments with a fluctuation amplitude less than the direction fluctuation threshold, and obtains a continuous stable direction sequence, where, represents the edge direction angle value of the th crack boundary pixel point, represents the total number of pixels participating in the direction angle change calculation in the current connected pixel segment, represents the direction angle change amplitude value in the th direction segment, represents the spatial spacing value of the pixel points in the th direction segment, represents the number of direction segments that can be classified in the current crack, represents the change gradient value of the th pixel point in the direction angle dimension, represents the number of participating points of the direction angle gradient within the current segment;

[0021] The path line construction sub-module extracts the spatial coordinate information of all pixel segments according to the continuous stable direction sequence, constructs path connection relationships based on the Euclidean distances and direction angle consistency values between adjacent coordinate points, and performs spatial path merging processing on continuous segments to obtain a set of crack main axis direction lines.

[0022] As a further solution of the present invention, the spatial encapsulation module includes:

[0023] The path projection sub-module extracts the set of spatial position points of all path lines in the building image coordinate system based on the set of crack main axis direction lines, constructs the projection relationship between the direction lines and the image reference coordinate axes according to the endpoint coordinates, direction vectors, and the included angles between continuous point pairs on each path line, and obtains the path spatial projection coordinate values;

[0024] The encapsulation area division sub-module extracts the start and end coordinates of each path line and constructs the circumscribed rectangular area of the line segment according to the path spatial projection coordinate values, calculates the included angle between the length direction of the circumscribed rectangle and the path direction, obtains the corrected angle value after direction alignment correction, and performs joint calculation in combination with the corrected angle value and the boundary coordinates of the circumscribed area using the formula:

[0025] ;

[0026] Calculate the encapsulation area range value of the th segment of the th crack , divide the encapsulation area according to the relative position of the encapsulation boundary in the image, establish an aligned encapsulation rectangle, and obtain a set of aligned encapsulation areas, where , are respectively the maximum and minimum abscissa values of the path line endpoints, , are the maximum and minimum ordinate values, , are respectively the included angles between the direction of the th segment of the path line and the horizontal and vertical axes, is the total length of the path segment in the encapsulation area, represents the total number of path segments;

[0027] The image block calibration sub-module extracts the position block numbers and boundary information of each encapsulation rectangle in the building image coordinate system according to the set of aligned encapsulation areas, performs number mapping on each encapsulation block according to the image row and column indexes, records the coordinate indexes and the corresponding path information of each encapsulation block, establishes a number list, and obtains a set of crack path encapsulation blocks.

[0028] As a further solution of the present invention, the structure overlap detection module includes:

[0029] The boundary data extraction sub-module encapsulates the block set based on the crack path, extracts the upper left and lower right coordinate points of each encapsulated block, calls the boundary coordinate set of the stress concentration area in the building structure image, arranges all boundary coordinate values in the format of a coordinate matrix, unifies the format, and then establishes a coordinate group to obtain the structural and block boundary coordinate set;

[0030] The spatial comparison calculation sub-module, according to the structural and block boundary coordinate set, conducts a spatial coordinate intersection detection on each encapsulated block and the building structure boundary area, compares whether there is a coverage relationship between the upper left and lower right coordinates, extracts the intersecting boundary points, and calculates the ratio of the intersecting area to the combined area, using the formula:

[0031] ;

[0032] Calculate the th encapsulated block and the th structural area overlap determination value , and compare it with the spatial coincidence range standard according to the determination value to identify the block numbers of all overlapping relationships, and obtain the overlapping block number set. Among them, represents the spatial overlapping area, represents the spatial union area, represents the length of the th intersection point connection line, represents the length of the corresponding direction line segment, represents the coordinate difference at the th position, are respectively the number of intersection point segments and the number of difference points, represents the two-dimensional spatial boundary area of the th crack path encapsulated block in the image coordinate system, represents the spatial boundary area of the th stress concentration area in the building structure image;

[0033] The overlapping number summary sub-module, according to the overlapping block number set, extracts all the encapsulated numbers with coordinate overlap with the stress area in the order of the encapsulated block numbers, injects the corresponding building image position index into the summary list, records the overlapping block number information, and obtains the structural block overlapping record table.

[0034] As a further solution of the present invention, the system further includes a building quality assessment module;

[0035] The building quality assessment module judges whether the cracks are concentrated in the key areas of the building structure according to the spatial coverage relationship between the number of overlapping blocks in the structural block overlapping record table and the total number of blocks in the crack path encapsulated block set, and divides different quality status categories to obtain the building structure quality assessment result;

[0036] The building structure quality assessment results include the crack distribution concentration index, the coverage rate of the key areas of the structure, and the quality grade classification results.

[0037] As a further solution of the present invention, the building quality assessment module includes:

[0038] The overlapping ratio calculation sub-module obtains the total number of encapsulated blocks and the number of blocks with overlapping relationships based on the structure block overlapping record table and the crack path encapsulated block set, counts the distribution quantity values of the two types of number sets in the building image coordinates, calculates the ratio of the overlapping quantity to the total number of encapsulated quantities, and obtains the spatial coverage value;

[0039] The distribution concentration determination sub-module performs coordinate matching on the positions of all encapsulated blocks with overlapping relationships in the building image according to the spatial coverage value, and combines the key stress area numbers and position coordinate ranges marked in the building structure diagram to determine whether they are concentrated in the key area, and obtains the key area concentration data;

[0040] The quality status division sub-module divides the building structure quality status categories based on the key area concentration data, taking the continuous block concentration value and the degree of spatial coincidence as the basis, generates the labels of each status level and the corresponding position indexes, and obtains the building structure quality assessment results.

[0041] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0042] In the present invention, through the directional analysis of the pixel gray-scale change of the crack boundary and the angle fluctuation constraint, a path line with stable and coherent characteristics is formed, enhancing the connectivity and continuity of the crack direction description. Using the spatial position relationship of the path line in the building image coordinates, an encapsulated rectangular block closely corresponding to the crack trend is generated, realizing the expansion of the crack expression from a linear path to a spatial block, improving the spatial organization expression ability. Each encapsulated block is then compared with the boundary information of the structural stress area at the coordinate level to form an overlapping expression between the defect influence area and the structural function area, enabling the defect recognition result to have the ability to judge the structural function association. Based on the quantitative coverage relationship between the total number of encapsulated blocks and the coincidence degree of the structural key area, combined with the image spatial distribution, a quality grade division mechanism is established to realize the comprehensive expression of the evaluation output for the crack spatial trend, geometric density and structural influence degree, making the defect expression move from graphic information to structural semantics. At the same time, the path continuity analysis and spatial encapsulation strategy are introduced to make the evaluation have multi-dimensional judgment ability, improving the detection coverage depth, the richness of the evaluation level and the accuracy of the building quality judgment. Description of the Drawings

[0043] Figure 1 is the system flow chart of the present invention;

[0044] Figure 2 This is the flowchart of the image acquisition module of the present invention;

[0045] Figure 3 This is the flowchart of the crack main axis identification module of the present invention;

[0046] Figure 4 This is the flowchart of the space encapsulation module of the present invention;

[0047] Figure 5 This is the flowchart of the structure overlap detection module of the present invention;

[0048] Figure 6 This is the flowchart of the building quality assessment module of the present invention. Specific implementation manners

[0049] In order to make the objectives, technical solutions and advantages of the present invention more clear and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0050] In the description of the present invention, it should be understood that the orientation or positional relationship indicated by the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the accompanying drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation of the present invention. In addition, in the description of the present invention, the meaning of "a plurality" is two or more, unless otherwise specifically defined.

[0051] Please refer to Figure 1 , the building engineering quality detection and evaluation system based on machine vision includes:

[0052] The image acquisition module acquires images of building surface structures such as walls, beams and columns, performs edge scanning on the non-background areas in the images, extracts the boundary pixels of the crack areas, and marks the positional relationship of each pixel in the building image coordinate system to obtain the crack boundary coordinate layer;

[0053] The crack main axis identification module calculates the edge direction angle value of each point according to the gray-scale change direction of the adjacent pixels of each crack boundary pixel point in the crack boundary coordinate layer, sorts the pixel point sequences according to the crack connection relationship, classifies the change trends of the direction angles in the sequences, and filters out the continuous pixel segments with a fluctuation amplitude less than the fluctuation threshold to construct the main axis trend path line, and obtains the crack main axis direction line set;

[0054] The spatial encapsulation module divides the encapsulation rectangular areas aligned with the crack trend based on the spatial position relationship of the path lines in the set of crack main axis direction lines on the building image coordinates, extracts the boundary coordinates of each area, records the positioning block numbers of each area in the building image, and obtains the crack path encapsulation block set;

[0055] The structure overlap detection module corresponds and compares the spatial boundary coordinates of each block in the crack path encapsulation block set with the boundary data of the stress concentration area marked in the building structure image to check whether there are spatial sections with overlapping coordinates, summarizes all the overlapping block numbers, and obtains the structure block overlap record table;

[0056] The building quality assessment module determines whether the cracks are concentrated in the key areas of the building structure based on the spatial coverage relationship between the number of overlapping blocks in the structure block overlap record table and the total number of blocks in the crack path encapsulation block set, and classifies different quality status categories to obtain the building structure quality assessment result.

[0057] The crack boundary coordinate layer includes the boundary pixel distribution map, the pixel coordinate relationship table, and the crack image position index. The set of crack main axis direction lines includes the main axis path line sequence, the direction angle sequence set, and the direction line classification label. The crack path encapsulation block set includes the encapsulation rectangular boundary data, the path block positioning number, and the block spatial trend information. The structure block overlap record table includes the overlapping block number table, the coincident space segment coordinate set, and the stress area correspondence mapping result. The building structure quality assessment result includes the crack distribution concentration index, the structure key area coverage rate, and the quality grade classification result.

[0058] Please refer to Figure 2 , the image acquisition module includes:

[0059] The image structure extraction sub-module acquires the building surface structure image, performs gray-scale conversion processing on the original image data, calculates and distinguishes the gray-scale difference between the background area and the non-background area, selects the background gray-scale threshold as the basis, classifies the area below the background gray-scale threshold as the non-background area, and obtains the non-background pixel distribution map;

[0060] The process of acquiring the building surface structure image usually uses a drone or a high-pixel image acquisition device as the preliminary equipment. During acquisition, it is necessary to ensure that the camera is perpendicular to the building surface and control the shooting distance within the range of 2 to 5 meters to reduce the distortion error. For the acquired image, it needs to be converted into a gray-scale image first to reduce the number of channels in the image processing process and improve the boundary recognition efficiency. The gray-scale conversion uses the weighted average method to process the RGB channel pixel values. The specific calculation method is: The grayscale value of each pixel = 0.299×R + 0.587×G + 0.114×B. Taking an image with RGB values of (120, 100, 90) as an example, the grayscale value is: 0.299×120 + 0.587×100 + 0.114×90 = 104.84;

[0061] After dividing the grayscale image into regions, read the pixel grayscale matrix, and compare the grayscale values of each pixel with the set background grayscale threshold in turn. The setting of the background grayscale threshold is based on the global grayscale histogram distribution of the image. Select the central value of the grayscale segment with the largest number of pixels as the background reference grayscale value. This value is generally set in the range of 180 - 200 according to the image statistical results. Select 190 as the background threshold (this setting refers to the grayscale distribution difference between the background area and the non-background area in the image. Especially for a concrete wall surface, the grayscale range formed under natural light conditions is concentrated in the interval of 180 - 210, while the grayscale of the areas corresponding to wall cracks and holes is mostly in the interval of 80 - 160. Therefore, 190 is selected as the background grayscale threshold. This value is in the middle region of the above grayscale intersection segment and can stably distinguish the concrete background area from the edge contour area. Under the conditions of image brightness enhancement or shadow enhancement, this value will linearly shift with the change of the overall pixel brightness mean value. It needs to be adjusted according to the grayscale mean value μg of the current image. Usually, within the interval μg ∈ [170, 200], the threshold 190 can meet the background filtering requirements of most images). Identify the area with a grayscale value lower than 190 as the non-background area, and filter out non-background dot matrices with a small size (less than 3×3 pixel blocks) to avoid noise interference. Finally, form a pixel distribution map of the non-background area, providing a pixel-based input for subsequent crack boundary recognition, as shown in Table 1.

[0062] Table 1 Example table for non-background area extraction

[0063]

[0064] As shown in Table 1, pixels with a grayscale value less than 190 are identified as non-background areas. After making this determination for all pixel points in the image, a non-background pixel distribution map can be obtained.

[0065] The edge pixel calibration sub-module, based on the non-background pixel distribution map, locates the edge contours of the non-background areas in the image, judges the gradient direction and the amplitude of grayscale change of adjacent pixel points in the image, marks the crack boundaries for pixels with an amplitude of grayscale change greater than the crack recognition gradient threshold, identifies all pixel coordinate information within the boundary contour line, and obtains the crack edge pixel marking result;

[0066] Based on the non-background pixel distribution map, it is necessary to further identify its edge contour. This process detects the amplitude of gray-scale change between pixels in an 8-neighborhood manner and calculates its directional gradient. The directional gradient takes the maximum value of the gray-scale difference between the current pixel point and the pixels to its right and below as the edge gradient value. For example, if the gray-scale value of point (25, 25) is 120, the pixel to its right (25, 26) is 160, and the pixel below (26, 25) is 110, then the gradient is max(|120 - 160|, |120 - 110|) = 40. Set the crack recognition gradient threshold to 35 (this value is obtained from the analysis of multiple typical crack images, and the gradient values in the crack area are mostly between 30 and 60). Therefore, the pixel point (25, 25) is marked as a crack boundary point. After performing this determination process pixel by pixel, all candidate pixel sets for crack edges are obtained. The candidate pixel points are arranged in the order of image coordinates and their coordinate values are recorded. Each coordinate point is stored in the form of (x, y), where x is the column coordinate and y is the row coordinate, forming preliminary crack contour marking data. On this basis, the part where the distance between consecutive pixels exceeds 2 pixels is removed to exclude non-connected crack areas. In an image with a resolution of 600×800, a total of 3200 crack edge pixels are identified, accounting for 2.5% of the non-background pixel quantity. Based on this, the crack edge pixel marking result is established.

[0067] The boundary coordinate generation sub-module, according to the crack edge pixel marking result, converts the positions of all pixel points within the crack boundary in the image matrix relative to the building image coordinate system, establishes the mapping relationship between the horizontal and vertical coordinate systems of the boundary points, and obtains the crack boundary coordinate layer;

[0068] According to the crack edge pixel marking result, convert the (x, y) image coordinates recorded in the marking value into the actual position coordinates in the building surface coordinate system. It is necessary to introduce the pixel spacing during image acquisition and the pixel ratio of the building surface for conversion. Assume the image resolution is 800×600, the actual covered width of the building is 4 meters, and the height is 3 meters. Then the physical size corresponding to each pixel is width: 4 / 800 = 0.005 meters, approximately 5 millimeters. Convert all pixel coordinates of the crack edge according to this ratio. For example, the pixel point (200, 300) is converted to (1.0 meter, 1.5 meters), which is the position of the crack in the building plane coordinate system. At the same time, establish a two-dimensional grid representation for all edge pixel points to form a crack boundary line network coordinate set. Transfer this coordinate set into the building image registration system and perform coordinate anchoring by comparing with the perspective view of the actual building drawing to complete the layer establishment work. The processing result of this process is the crack boundary coordinate layer.

[0069] Please refer to Figure 3 , the crack main axis recognition module includes:

[0070] The direction angle extraction sub-module calculates the difference between the gray values of adjacent pixels and the target pixel for each crack boundary pixel in the crack boundary coordinate layer, computes the distribution mean and change rate of the gray difference in the adjacent pixel directions, determines the gray change direction, measures the gradient angle and converts the direction vector for each crack boundary pixel, and obtains the edge direction angle value;

[0071] Based on each crack boundary pixel in the crack boundary coordinate layer, the gray values of the pixel to be recognized in the image and its 8 adjacent direction pixels are obtained. After performing the difference operation on each pair of gray values, direction judgment is carried out. Among them, the pixel point is set as and its 8 adjacent pixels in different directions are respectively denoted as If the gray value of the pixel point is 128, and the gray values of the adjacent points are 131, 129, 124, 120, 122, 130, 135, 127 respectively, the direction change is mainly located in the direction areas 2 to 5 of the gray decline trend, indicating that there is an edge direction vector at this point. Then, the gradient vector in this direction range is converted into an angle value, and the gray change direction is converted into an angle range of 0° to 180° using the arctangent operation. In this example, the gradient change corresponding to the directions 2 to 5 is negative, and the converted angle is about 110°. This value is recorded as the direction angle of this pixel. The above operations are performed on all crack boundary pixels in turn to form an angle map matrix. The range of the direction angle values is usually distributed between 5° and 175° according to the image resolution and sampling interval, and the data is saved in integer form. If there is a discontinuous gray change in the boundary transition area of the image area, the direction angles of the pixel points with a difference less than 3 are temporarily set as invalid values and removed in the next stage; during this process, it is necessary to perform interval determination on the range of the direction angle values, and set the gray difference threshold as ±5. If the gray differences of the adjacent directions of a certain pixel point are all less than this value, it is determined that this point does not have direction angle information and is recorded as 0°. For example, the gray value of point B is 126, and the maximum difference of its adjacent points is 3, then its direction angle is set as 0°. After completing the angle recognition of all boundary points, the edge direction angle value is generated.

[0072] The sequence trend classification sub-module analyzes the connectivity relationship of the crack boundary pixels according to the edge direction angle values, sorts the sequences of connected pixels in adjacent order, calculates the direction angle change rate of each pixel point and the difference between the direction angles of adjacent points, classifies the direction angle change trends, and uses the formula:

[0073] ;

[0074] The operation obtains the direction fluctuation trend value , filters out the continuous pixel segments with a fluctuation amplitude less than the direction fluctuation threshold, and obtains a continuous stable direction sequence. Among them, represents the edge direction angle value of the th crack boundary pixel point, Indicates the total number of pixels participating in the calculation of the direction angle change in the current connected pixel segment, Indicates the magnitude of the direction angle change in the th direction segment, Indicates the spatial spacing value of the pixel points in the th direction segment, Indicates the number of direction segments that can be classified in the current crack, Indicates the change gradient value of the th pixel point in the direction angle dimension;

[0075] According to the edge direction angle value, call the connectivity relationship of the crack boundary pixel points, extract the pixel sequences of each connected segment, sort them by horizontal and vertical coordinates to form an ordered pixel chain. In a certain crack segment, assume that the extracted pixel sequence has 15 points, and each point has a corresponding angle value, such as 78°, 79°, 80°, 81°, 130°, 133°, 135°, 137°, 140°, 138°, 139°, 141°, 82°, 80°, 79°. First, calculate the difference in the direction angles of adjacent points. If the angle difference in a certain segment does not exceed 5°, it is recorded as a stable direction segment. Accumulate the angle differences within this segment to find the total fluctuation amount, and at the same time multiply it by the pixel distance between its corresponding points, perform a sum of squares operation, and then divide by the square root of the sum of the absolute values of the direction angle gradient values within this segment to calculate the direction fluctuation trend value. This process uses the formula:

[0076] ;

[0077] Among them, the data for each segment is as follows:

[0078] : 78, 79, 80, 81, 130, 133, 135, 137, 140, 138, 139, 141, 82, 80, 79 in sequence;

[0079] : The angle differences for each segment are 1, 1, 50, 3, 3 respectively;

[0080] : The average pixel spacing within the segment is set to 1.2 pixels;

[0081] : The direction gradient values are 2, 2, 4, 3, 6, 5, 4, 4, 3, 2, 2, 3;

[0082] Substituting gives:

[0083] ;

[0084] The calculated direction fluctuation trend value for this crack segment is 775.92. The set fluctuation threshold is 800 (the basis for its setting is that the angular variation range of the common stable main axis segment in the crack image usually focuses on an angular amplitude change not exceeding 8°, and the average gradient intensity is in the range of 40 to 60. Combining the image sampling density (set to 1.2 pixels / mm in this image) and the crack length standard segment set between 12 and 20 pixels, then the average direction fluctuation trend value calculated in multiple consecutive segment samples is between 700 and 850. The fluctuation threshold of 800 is the corresponding value of the 85th percentile in 40 samples as the judgment standard, and this value shows a positive correlation trend as the image resolution increases or the crack complexity improves, and it needs to be dynamically adjusted in combination with the pixel spacing, total crack length, and angular variability), and a value lower than this indicates that this segment is a direction continuous segment, which is recorded in the continuous stable direction sequence value.

[0085] The operation logic of the formula is based on the comprehensive measurement of the direction angle change trend of the crack boundary pixels. The numerator part sums the absolute values of the direction angle differences between adjacent pixel points, reflecting the cumulative fluctuation of the overall direction change in the pixel sequence. Subsequently, the angular change amplitude of each direction segment is multiplied by its corresponding pixel spacing, representing the comprehensive weight of the direction fluctuation of this segment on the spatial scale, and then squared to emphasize the dominant role of the large-amplitude continuous fluctuation segment in the overall trend value; while the denominator part sums the absolute values of the direction gradient change values of all participating points and then takes the square root to form a normalization suppression term for the fluctuation trend. The logic is that if the direction angle of a crack segment changes violently, but at the same time the gradient intensity is also high, the influence of the direction fluctuation of this segment in the overall trend is weakened, thus preventing the interference of the overall trend judgment due to a single-point sharp change. Finally, this formula constructs a fractional relationship with the "total fluctuation + square term of the spatial influence weight" as the numerator and the "square root of the total direction consistency gradient" as the denominator, forming a comprehensive evaluation index for the coherence of the main axis direction.

[0086] The path line construction sub-module extracts the spatial coordinate information of all pixel segments according to the continuous stable direction sequence, constructs a path connection relationship based on the Euclidean distance and direction angle consistency value between adjacent coordinate points, and performs spatial path merging processing on the continuous segments to obtain the set of crack main axis direction lines;

[0087] According to the continuous stable direction sequence, the spatial coordinate values of each pixel point in each segment are selected. Suppose a stable pixel direction sequence in a certain segment contains pixel points A(50, 80), B(51, 81), C(52, 82), D(53, 83), E(54, 84). Then the distances between each point are extracted in turn, and the spatial distances between adjacent points are calculated through the Euclidean formula. After connecting them in turn, a set of line segments is formed. In this process, a direction angle consistency index is set. When the average difference of the direction angles of adjacent three points does not exceed the set value of 5°, they are merged into one path segment. Finally, multiple path segments are merged into a main axis set, which is stored in the form of an array, recording the starting point, ending point and the average value of the path direction angle of each main axis. The structural example is as follows:

[0088] Table 2 Parameter Table of Main Axis Direction Lines

[0089]

[0090] As shown in Table 2, the path of main axis No. 1 is composed of 5 consecutive coordinate points, the direction angle is concentrated between 80° and 82°, the path length is 6.4 pixels, which is used to identify the main crack trend in the subsequent crack structure recognition, and finally a set of main crack axis direction lines is generated.

[0091] Please refer to Figure 4 , the space encapsulation module includes:

[0092] Based on the set of main crack axis direction lines, the path projection sub-module extracts the set of spatial position points of all path lines in the building image coordinate system, constructs the projection relationship between the direction line and the image reference coordinate axes according to the endpoint coordinates, direction vectors and the included angles between consecutive point pairs on each path line, and obtains the path space projection coordinate values;

[0093] Based on the spatial position relationship of the path lines in the building image coordinates in the set of main crack axis direction lines, first extract all the endpoint coordinates of each path in the path line set and the horizontal and vertical distribution coordinate values of the path continuous points. For example, select a path segment identified in an image, its starting coordinate is point A(150, 240), and the ending coordinate is point B(320, 410). Then this path line can be defined as a vector set, composed of several pixel pairs. Subsequently, according to the horizontal coordinate range and the vertical coordinate range , calculate its projection boundary in the image, record each pixel point on the path line by point-by-point scanning, and record its horizontal and vertical index values in the image matrix. In the processing process, to ensure the coordinate projection accuracy, taking a building image with a pixel resolution of 0.25mm / px as an example, if the horizontal span of the path line is 680px, then its corresponding projection length is 170mm. At the same time, combined with the direction vector of each path line, such as the direction vector of the path line is , the angle between it and the horizontal axis of the building image can be calculated through the inverse trigonometric function , finally, this angle is used to determine the projection relationship between the path direction and the image reference coordinates, extract the direction angle of each crack path line, and accordingly adjust the coordinate boundary range. At the same time, bind and encode the path line number with its corresponding coordinate value and output it to obtain the path space projection coordinate value.

[0094] The encapsulation area division sub-module extracts the start and end coordinates of each path line according to the path space projection coordinate value, constructs the circumscribed rectangle area of the line segment, calculates the angle between the length direction of the circumscribed rectangle and the path direction, and obtains the correction angle value after direction alignment correction. Combine the correction angle value with the boundary coordinates of the circumscribed area for joint calculation, using the formula:

[0095] ;

[0096] Calculate the th crack range value of the encapsulation area of the th path segment, divide the encapsulation area according to the relative position of the encapsulation boundary in the image, establish an aligned encapsulation rectangle, and obtain the set of aligned encapsulation areas, where 、 are the maximum and minimum abscissa values of the path line endpoints respectively, 、 are the maximum and minimum ordinate values, 、 are the angles between the th path line direction and the horizontal axis and vertical axis respectively, is the total length of the path segment in the encapsulation area, represents the total number of path segments;

[0097] There is an inclusion and constitution relationship between the crack and the path, that is, each crack is composed of multiple path segments, and the path is the spatial expression form of the crack. When extracting cracks in the image, the crack will be divided into several path segments, each path segment has clear start and end coordinates and direction angles, and the encapsulation area is constructed based on this.

[0098] According to the path space projection coordinate value, extract the coordinate boundaries of the start and end points of each path segment in the building image, calculate the boundary coordinate range of the circumscribed rectangle required for encapsulating the path, and perform angle alignment processing on the rectangle and the path line direction. Use the angle correction method to reset the rectangle boundary direction. For example, the minimum and maximum abscissas of the path line in the image are 、 , the minimum and maximum ordinates are 、 , then its width and height are 170px and 170px respectively. Combine the angle between this path line and the horizontal axis, the vertical axis angle , the number of direction segments is 3, and the total path length is , substitute into the formula:

[0099] ;

[0100] Each angle term is respectively , , and take their cosine and sine values respectively as:

[0101] ;

[0102] ;

[0103] ;

[0104] Substitute the above data into:

[0105] ;

[0106] Table 3 Crack Path Line Encapsulation Parameter Table:

[0107]

[0108] As shown in Table 3, the encapsulation area range value is used to calibrate the comprehensive index of the encapsulation boundary, and is evaluated in combination with the direction projection angle, the boundary coordinate range, and the total path length.

[0109] The operation logic of this formula lies in comprehensively measuring the spatial range and direction consistency of the path encapsulation area. Specifically, the in the formula reflects the sum of the horizontal and vertical dimensions of the encapsulation rectangle in the image coordinate system, that is, the overall scale of the encapsulation boundary, as the measurement basis for the degree of path space occupation; and the in the numerator part represents the sum of the cosine and sine values of the angle between the path line direction and the coordinate axis, used to comprehensively evaluate the coordination degree of the path direction distribution on the two axes in the image coordinate system, reflecting the matching degree between the path line direction and the image encapsulation direction. The greater this direction matching degree, the better the alignment of the encapsulation path; the in the denominator is used to normalize the encapsulation index of the entire path segment, is the total length of the path segment. After taking the square root, it compresses the weight influence of the extreme length segments, and the +1 operation avoids the abnormal situation of the denominator being 0, and at the same time buffers the ratio amplification caused by the length. The overall logic is based on the actual geometric size of the path encapsulation area, supplemented by the component contribution of the path direction and corrected for the total path length, so that the obtained encapsulation area range value achieves a balanced representation among the spatial scale, direction alignment, and length control.

[0110] The image block calibrating sub-module extracts the position block numbers and boundary information of each encapsulation rectangle in the building image coordinate system according to the alignment encapsulation area set, performs number mapping on each encapsulation block according to the image row and column indexes, records the coordinate index and the path information where each encapsulation block is located, establishes a number list, and obtains the crack path encapsulation block set;

[0111] According to the alignment encapsulation area set, extract the rectangular boundary coordinates of each encapsulation area, locate the image block numbers to which each encapsulation area belongs according to the horizontal and vertical index values in the building image matrix, and perform number mapping processing on the image partition range. If the image size is 1920×1080px and is divided into one block every 320×270px, then the whole image is divided into 24 blocks in total of 6×4. For example, if the boundary of a certain encapsulation rectangle area is the horizontal coordinate interval 190–290 and the vertical coordinate interval is 300–520, then calculate that the block number is the first block on the horizontal axis and the second block on the vertical axis, and the block number is (1, 2). The corresponding number ID can be block number ID = 1×4 + 2 = 6. Perform similar mapping records on all encapsulation areas. Finally, establish the building image block number index table corresponding to each encapsulation path, establish the mapping relationship, and output the three items of information of the corresponding encapsulation path, boundary coordinates, and block numbers, that is, obtain the crack path encapsulation block set.

[0112] Please refer to Figure 5 , the structure overlap detection module includes:

[0113] The boundary data extraction sub-module extracts the upper left and lower right coordinate points of each encapsulation block based on the crack path encapsulation block set, calls the boundary coordinate set of the stress concentration area in the building structure image, arranges all boundary coordinate values in the coordinate matrix format, and establishes a coordinate group after unifying the format to obtain the structure and block boundary coordinate set;

[0114] Based on the crack path encapsulation block set, extract the upper left and lower right coordinate points of each encapsulation block, perform normalization processing on the coordinate values in the image coordinate axis direction, and unify the pixel unit to the millimeter unit for scale-consistent spatial overlap judgment. Extract the boundary coordinates of the stress concentration area in the building structure image. The boundary information is expressed by linear markings or area frame lines in the structure drawing. In the example, it is assumed that the upper left coordinate of the encapsulation block is (120mm, 80mm), the lower right coordinate is (180mm, 140mm), the upper left coordinate of the structure area boundary is (160mm, 100mm), and the lower right coordinate is (220mm, 160mm). The two groups of boundary values are unified into arrays respectively as and , detect whether there is an intersection relationship between the two groups of boundaries, and perform interval intersection judgment on the boundary ranges. First, judge whether there is an intersection in the x-axis direction, that is, the maximum left boundary is less than the minimum right boundary, It holds, indicating that there is an overlapping area in the x direction, and judgment is made in the y direction It also holds. Therefore, there is an intersection area between the two in space. Record the intersection coordinates as (160, 100) to (180, 140). This part of the area can be called by subsequent modules for overlapping area calculation, and finally establish the set of structural and block boundary coordinates.

[0115] According to the set of structural and block boundary coordinates, the spatial coordinate intersection detection is carried out for each encapsulated block and the building structure boundary area. Compare whether there is a covering relationship between the upper left and lower right coordinates, extract the intersection boundary points, and calculate the ratio of the intersection area to the union area. The formula is used:

[0116] ;

[0117] Calculate the th overlapping determination value of the encapsulated block and the th structural area , and compare it with the spatial coincidence range standard according to the determination value to identify the serial numbers of all blocks with coincidence relationships, and obtain the set of overlapping block numbers. Among them, represents the spatial overlapping area, represents the spatial union area, represents the length of the intersection point connection line of the th segment, represents the length of the corresponding direction line segment, represents the coordinate difference at the th position, are the number of intersection segments and the number of difference points respectively, represents the two-dimensional spatial boundary area of the th crack path encapsulated block in the image coordinate system, represents the spatial boundary area of the stress concentration area in the th building structure image;

[0118] According to the set of structural and block boundary coordinates, extract the intersection coordinates corresponding to the encapsulated block and the structural area, perform area calculation and comparison with the union area. The overlapping area is the product of the width and height of the intersection rectangle. In the above example, the width of the intersection rectangle is , the height is , so the overlapping area is , the total area of the encapsulated block is , the area of the structural area is , the union area is , substitute the above values into the formula, where the length of the boundary intersection connection line , the length of the corresponding direction line segment , the coordinate difference , after substitution, we get:

[0119] ;

[0120] The result shows that the overlapping degree is about 0.5. The ratio threshold is set to 0.3 in the overlapping determination standard (the setting basis is that the average proportion of the stress concentration area in the building structure image in the total pixel area of the image is between 24.6% and 32.8%. To consider the maximum redundant coincidence impact of the path encapsulation block in the case of extreme overlap, it is set that the overlap rate higher than the threshold near the upper bound may have an interference impact on the structure. Therefore, 0.3 is used as the risk identification boundary value, which fluctuates with the expansion degree of the stress area boundary and the image resolution in the structure annotation map. When the boundary range expands or the resolution decreases, the overlapping area will increase, and then the overlapping ratio will increase. Therefore, this threshold should be adjusted synchronously according to the physical scale standard of the image and the granularity of the structure area division). If the overlapping determination value exceeds this standard, it is identified as a structural coincidence block. Then the current result meets the conditions and should be included in the coincidence record range.

[0121] Table 4 Overlap calculation table of the encapsulation area and the structure area

[0122]

[0123] Table 4 lists the actual calculation process and results of the spatial overlap, verifying that the formula can be used to identify the structural overlap area.

[0124] The operation logic of the formula is to quantitatively measure the spatial relationship between the crack encapsulation block and the structure area in the image coordinate system, so as to comprehensively reflect their spatial coincidence intensity and geometric feature coupling situation. The numerator part of the formula consists of two items. The first item is the overlapping area , which directly reflects the overlapping range of the encapsulation area and the structure area in space. The second item is the sum of the products of the lengths of all overlapping boundary line segments and their direction line segments, which reflects the along - direction consistency and continuity of the overlapping contour, and reflects the geometric similarity between the crack path and the structure boundary. The sum of the two items gives the total intensity index of the structural overlap; the first item in the denominator part is the combined area of the encapsulation area and the structure area, which is used to normalize the overlapping intensity. The second item is the square root sum of the absolute values of the coordinate differences of all pixels in the overlapping area, which is intended to reflect the geometric distortion degree of the junction area. The square root operation enhances the sensitivity to small - amplitude boundary changes, and finally forms a composite ratio with boundary coincidence degree, geometric consistency and normalization control, which can present the coincidence relationship between the encapsulation block and the structure area in multiple dimensions.

[0125] The overlapping number summary sub - module extracts all encapsulation numbers that have coordinate coincidence with the stress area in the order of the encapsulation block serial number according to the set of overlapping block numbers, injects the corresponding building image position index into the summary list, records the overlapping block number information, and obtains the structural block overlapping record table;

[0126] According to the set of overlapping block numbers, screen all block numbers that meet the condition that the overlapping determination value is greater than the standard value (such as 0.3), organize their numbers, positions, and image row-column index information, and establish an index mapping table. In the example, A1 that meets the condition is used as a valid record item. The upper-left coordinate of it in the image is located at (120, 80). In the case of an image resolution of 1mm / pixel, the corresponding row and column numbers are the 80th row and the 120th column. Therefore, the record number is , the structure area number is , the position index is (120, 80), and finally output a tabular structure block overlap record set. This table can provide a spatial correspondence reference for structural response analysis or image risk detection, and finally obtain a structural block overlap record table.

[0127] Please refer to Figure 6 , the building quality assessment module includes:

[0128] The overlap ratio calculation sub-module, based on the structural block overlap record table and the crack path encapsulated block set, obtains the total number of encapsulated blocks and the number of blocks with an overlapping relationship, counts the distribution quantity values of the two types of number sets in the building image coordinates, calculates the ratio of the overlapping number to the total number of encapsulated blocks, and obtains the spatial coverage value;

[0129] Based on the structural block overlap record table and the crack path encapsulated block set, first extract the marked block numbers in the structural block overlap record table and store them as a number array , and extract all block numbers from the encapsulated block set to form a complete set , count the numbers in these two sets correspondingly, count the total number of overlapping numbers and the total number of encapsulated blocks, and obtain the overlap ratio = overlapping encapsulated blocks / total number of encapsulated blocks. If the total number of extracted encapsulated blocks in the actual project is 50 and the number of overlapping encapsulated blocks detected with the structural blocks is 12, then the overlap ratio = 12÷50 = 0.24. This is, subsequently, for each overlapping block, extract its encapsulated area. The coordinates of the encapsulated area are represented by the upper-left vertex and the lower-right vertex to represent, and use the area formula to calculate the area of each overlapping block. If the coordinates are from (100, 120) to (180, 200), then A = (180 - 100)×(200 - 120) = 80×80 = 6400 pixel². At the same time, the corresponding area of the encapsulated block complete set is also calculated. For example, another block The area of all overlapping blocks and the area of the encapsulated set is (50, 60) to (130, 140), and its area is 6400 pixels². The statistical areas of all overlapping blocks and the encapsulated set are ΣX and ΣY respectively. In the example, if ΣX=76800 pixels², ΣY=256000 pixels², then the area ratio is 76800÷256000=0.3, and a spatial offset statistical set is constructed. The center coordinates of the overlapping blocks are further identified on the structural diagram, and the distance difference with the geometric center point of the corresponding structural stress area is calculated and statistically analyzed. Combined with all the above data, a comprehensive comparison is performed to obtain the spatial coverage of the influence range of the building cracks. The process is shown in Table 5:

[0130] Table 5 Statistics of building encapsulation blocks

[0131]

[0132] As shown in Table 5, some numbers in the package block have spatial overlap. The spatial coverage value can be directly obtained by calculating the ratio of the overlapping area to the total package area and comparing the number statistics.

[0133] The distribution concentration determination submodule matches the coordinates of all the overlapping encapsulated blocks in the building image according to the spatial coverage value, and determines whether they are concentrated in the key area by combining the key stress area number and position coordinate range marked in the building structure diagram, and obtains the key area concentration data;

[0134] According to the spatial coverage value, each overlapping encapsulated block is positioned as a point set in the image coordinate system {( , ), , …, }, and synchronously call the key structural area position set set in the building structure image { , , …, }, by judging the relationship between the coordinate points, check whether the packaging area falls into the above key structure area enclosing box one by one. For example, if the packaging block number The center coordinates are (140, 160), and the key area The boundary of is from (120, 140) to (200, 220), then Fall into the key structural area, marked as "concentrated". In specific application scenarios, for example, for the structure of a five-story office building, the key stress area is the interface between the concentrated positions of load-bearing columns and shear walls. Cracks encapsulation blocks are densely aggregated at the boundary of this area in the images of the third and fourth floors. The system further counts the number of concentrated encapsulation areas and their proportion in the total number of overlapping blocks. If there are 10 overlapping blocks and 7 of them are located in the key structural area, the concentration value is 0.7. According to the actual engineering setting, when the concentration ≥ 0.6, it is determined as "concentrated distribution", and if the concentration < 0.3, it is determined as "non-concentrated". This division standard is set with reference to the importance level of building structures and safety assessment indicators. A mapping set is formed between the concentration value and the key area number, and then the key area concentration data is obtained.

[0135] The sub-module for dividing the quality status divides the quality status categories of the building structure based on the key area concentration data, taking the concentration value of continuous blocks and the degree of spatial coincidence as the basis, generates labels for each status level and the corresponding position indexes, and obtains the building structure quality assessment result;

[0136] According to the key area concentration data, compare this value with the preset reference value for dividing the building quality status levels. It is set that the building structure quality levels are divided into five categories, and the corresponding concentration value intervals are: Category I [0, 0.2), Category II [0.2, 0.4), Category III [0.4, 0.6), Category IV [0.6, 0.8), Category V [0.8, 1]. For example, if the previously calculated concentration is 0.7, then the building structure assessment level is Category IV, and its position range and distribution number are marked. A two-dimensional mapping structure is constructed in the system output, with the image position coordinates as the index and the quality level as the value, forming a spatial distribution quality assessment matrix. In addition, for those with unclear boundary coincidence but in important structural functional areas (such as the beam end interface, stair connection parts, etc.), supplementary judgment is also required. If the crack sequence has continuous overlapping blocks exceeding the set lower limit (such as ≥ 3), it can be forced to be classified as the concentrated category to ensure the comprehensiveness of the assessment determination. Record the final divided level and its corresponding encapsulation block number, spatial coordinates, concentration degree and other elements into the output structure table as the basis for subsequent structural safety analysis, and obtain the building structure quality assessment result.

[0137] The above is only the preferred embodiment of the present invention, and it does not limit the present invention in other forms. Any person skilled in the art may use the disclosed technical content to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, as long as it does not depart from the technical solution content of the present invention, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention still belong to the protection scope of the technical solution of the present invention.

Claims

1. A machine vision-based building engineering quality inspection and evaluation system, characterized in that, The system includes: The image acquisition module acquires the building surface structure image, performs edge scanning on the non-background area of the image, extracts the boundary pixels of the crack area, and marks the position relationship of each pixel in the building image coordinate system to obtain the crack boundary coordinate layer; The crack main axis recognition module calculates the edge direction angle value of the point according to the gray-scale change direction of the adjacent pixels of each crack boundary pixel point in the crack boundary coordinate layer, sorts the pixel point sequence and classifies the change trend of the direction angle, constructs the main axis trend path line, and obtains the crack main axis direction line set; The space encapsulation module divides the encapsulation rectangular area aligned with the crack trend based on the spatial position relationship of the path lines in the crack main axis direction line set on the building image coordinates, extracts the boundary coordinates of each area, and records the positioning block numbers of each area to obtain the crack path encapsulation block set; The structure overlap detection module compares the spatial boundary coordinates of each block in the crack path encapsulation block set with the boundary data of the stress concentration area in the building structure image to check whether there is a spatial section with overlapping coordinates, summarizes the overlapping block numbers, and obtains the structure block overlap record table.

2. The machine vision-based building engineering quality inspection and evaluation system according to claim 1, characterized in that, The crack boundary coordinate layer includes the boundary pixel distribution map, the pixel coordinate relationship table, and the crack image position index. The crack main axis direction line set includes the main axis path line sequence, the direction angle sequence set, and the direction line classification label. The crack path encapsulation block set includes the encapsulation rectangular boundary data, the path block positioning number, and the block spatial trend information. The structure block overlap record table includes the overlapping block number table, the overlapping space segment coordinate set, and the stress area correspondence mapping result.

3. The machine vision-based building engineering quality inspection and evaluation system according to claim 1, characterized in that, The image acquisition module includes: The image structure extraction sub-module acquires the building surface structure image, performs gray-scale conversion processing on the original image data, calculates and distinguishes the gray-scale difference between the background area and the non-background area, selects the background gray-scale threshold as the basis, classifies the area below the background gray-scale threshold as the non-background area, and obtains the non-background pixel distribution map; The edge pixel calibration sub-module locates the edge contour of the non-background area in the image based on the non-background pixel distribution map, judges the gradient direction and gray-scale change amplitude of adjacent pixel points in the image, marks the crack boundary for the pixels with a gray-scale change amplitude greater than the crack recognition gradient threshold, and identifies all pixel coordinate information within the boundary contour line to obtain the crack edge pixel marking result; The boundary coordinate generation sub-module converts the positions of all pixel points within the crack boundary in the image matrix relative to the building image coordinate system according to the crack edge pixel marking result, establishes the mapping relationship between the horizontal and vertical coordinate systems of the boundary points, and obtains the crack boundary coordinate layer.

4. The machine vision-based building engineering quality inspection and evaluation system according to claim 1, characterized in that, The crack main axis recognition module includes: The direction angle extraction sub-module obtains the difference between the gray-scale value of the adjacent pixel and the gray-scale value of the target pixel based on each crack boundary pixel point in the crack boundary coordinate layer, calculates the distribution mean and change rate of the gray-scale difference in the adjacent pixel direction, judges the gray-scale change direction, measures the gradient angle and converts the direction vector for each crack boundary pixel point to obtain the edge direction angle value; The sequence trend classification sub-module analyzes the connectivity relationship of crack boundary pixel points according to the edge direction angle value, sorts each connected pixel point sequence in adjacent order, calculates the direction angle change rate of each pixel point and the difference in direction angles between adjacent points, classifies the direction angle change trend, and uses the formula: ； Obtain the directional fluctuation trend value through calculation , filter out consecutive pixel segments with a fluctuation amplitude smaller than the directional fluctuation threshold to obtain a continuous stable direction sequence, where represents the edge direction angle value of the th crack boundary pixel point, represents the total number of pixels participating in the calculation of the directional angle change in the current connected pixel segment, represents the directional angle change amplitude value in the th direction segment, represents the spatial spacing value of the pixel points in the th direction segment, represents the number of direction segments that can be classified in the current crack, represents the change gradient value of the th pixel point in the dimension of the directional angle, represents the number of participating points of the directional angle gradient within the current segment; The path line construction sub-module extracts the spatial coordinate information of all pixel segments in sequence order according to the continuous and stable direction sequence, constructs a path connection relationship based on the Euclidean distance between adjacent coordinate points and the direction angle consistency value, and performs spatial path merging processing on continuous segments to obtain a set of crack main axis direction lines.

5. The machine vision-based building engineering quality inspection and evaluation system according to claim 1, wherein The spatial encapsulation module includes: The path projection sub-module extracts the set of spatial position points of all path lines in the building image coordinate system based on the set of crack main axis direction lines, constructs the projection relationship between the direction line and the image reference coordinate axis according to the endpoint coordinates, direction vectors, and angles between continuous point pairs on each path line, and obtains the path space projection coordinate values; The encapsulation area division sub-module extracts the start and end coordinates of each path line and constructs the circumscribed rectangle area of the line segment according to the path space projection coordinate values, calculates the angle between the length direction of the circumscribed rectangle and the path direction, obtains the correction angle value after direction alignment correction, and performs joint calculation in combination with the correction angle value and the boundary coordinates of the circumscribed area, using the formula: ； Calculate the encapsulation area range value of the segment path of the th crack. Divide the encapsulation area according to the relative position of the encapsulation boundary in the image, establish an encapsulation rectangle with the same direction, and obtain a set of aligned encapsulation areas. Among them, and are the maximum and minimum abscissa values of the path line endpoints respectively, and are the maximum and minimum ordinate values, and are the angles between the th segment path line and the horizontal and vertical axes respectively, is the total length of the path segments in the encapsulation area, represents the total number of path segments; The image block calibration sub-module extracts the position block number and boundary information of each encapsulation rectangle in the building image coordinate system according to the set of aligned encapsulation areas, performs number mapping on each encapsulation block according to the image row and column indexes, records the coordinate index and the path information corresponding to each encapsulation block, establishes a number list, and obtains the crack path encapsulation block set.

6. The machine vision-based construction project quality inspection and evaluation system according to claim 1, characterized in that, The structure overlap detection module includes: The boundary data extraction sub-module extracts the upper left and lower right coordinate points of each encapsulation block based on the set of crack path encapsulation blocks, calls the set of boundary coordinates of the stress concentration area in the building structure image, arranges all boundary coordinate values in a coordinate matrix format, and establishes a coordinate group after unifying the format to obtain the set of structure and block boundary coordinates; The spatial comparison calculation sub-module performs spatial coordinate intersection detection on each encapsulation block and the building structure boundary area according to the set of structure and block boundary coordinates, compares whether there is a coverage relationship between the upper left and lower right coordinates, extracts the intersecting boundary points, and calculates the ratio of the intersecting area to the combined area, using the formula: ； Calculate the overlap determination value between the th encapsulated block and the th structural area. Compare the determination value with the spatial coincidence range standard to identify the serial numbers of all blocks with a coincidence relationship, and obtain the set of overlapping block numbers. Among them, represents the spatial overlapping area, represents the spatial union area, represents the length of the intersection connection line of the th segment, represents the length of the line segment in the corresponding direction, represents the coordinate difference at the th position, are respectively the number of intersection segments and the number of difference points, represents the two-dimensional spatial boundary region of the th crack path encapsulated block in the image coordinate system, represents the spatial boundary region of the stress concentration area in the th building structure image; The overlapping number summary sub-module extracts all the encapsulation numbers that have coordinate overlaps with the stress area in sequence order according to the set of overlapping block numbers, injects the corresponding building image position indexes into the summary list, records the overlapping block number information, and obtains the structure block overlap record table.

7. The machine vision-based building engineering quality inspection and evaluation system according to claim 1, characterized in that The system further includes a building quality assessment module; The building quality assessment module determines whether the cracks are concentrated in the key areas of the building structure according to the spatial coverage relationship between the number of overlapping blocks in the structure block overlap record table and the total number of blocks in the set of crack path encapsulation blocks, and divides different quality status categories to obtain the building structure quality assessment result; The building structure quality assessment results include the crack distribution concentration index, the coverage rate of the key areas of the structure, and the quality grade classification results.

8. The machine vision-based building engineering quality inspection and evaluation system according to claim 7, wherein, The building quality assessment module includes: The overlapping ratio calculation sub-module obtains the total number of encapsulated blocks and the number of blocks with overlapping relationships based on the structure block overlapping record table and the crack path encapsulated block set, counts the distribution quantity values of the two types of number sets in the building image coordinates, calculates the ratio of the overlapping quantity to the total number of encapsulated quantities, and obtains the spatial coverage value; The distribution concentration determination sub-module performs coordinate matching on the positions of all encapsulated blocks with overlapping relationships in the building image according to the spatial coverage value, combines the key stress area numbers and position coordinate ranges marked in the building structure diagram, and determines whether they are concentrated in the key area to obtain the key area concentration data; The quality status division sub-module divides the building structure quality status categories based on the key area concentration data, taking the continuous block concentration value and the degree of spatial coincidence as the basis, generates the labels of each status level and the corresponding position index, and obtains the building structure quality assessment results.

Images