Slope support progress calculation method based on image processing

Abstract

The invention relates to the technical field of slope engineering management and computer vision, and discloses a slope support progress calculation method based on image processing. The method comprises the following steps: firstly, obtaining a slope support field picture with a timestamp According to a feature point identification criterion, through an image identification technology, obtaining feature points in the on-site picture; correcting the on-site picture into an on-site picture in an orthogonal view form through homography transformation by taking the feature points in the on-site picture as base points and combining with a slope support design drawing; in the field picture in the orthogonal view form, obtaining the area of the slope supporting structure through a similarity algorithm; and finally, calculating the slope supporting progress based on the timestamp and the area of the slope supporting structure. According to the method, the support progress change is efficiently, flexibly and dynamically captured, and the response speed of engineering management is improved.

Description

Technical Field

[0001] The present invention relates to the fields of slope engineering management and computer vision technology, and in particular, to a method for calculating the progress of slope support based on image processing. Background Art

[0002] During the construction process of slope engineering, the progress monitoring of the support structure is the core link to ensure the safety and quality of the project. Traditional monitoring methods mainly rely on professional equipment such as total stations and laser scanners, and obtain parameters such as the scope and area of the support structure through manual fixed-point measurement. For example, construction workers need to regularly carry equipment into high-risk slope areas, collect data point by point and record it manually. The single monitoring cycle often takes several hours or even longer. This method not only has a significant pressure on labor costs, but also has a high data collection error rate due to environmental factors such as bad weather and complex terrain. Especially under extreme climate conditions such as flood seasons or typhoons, the feasibility of manual inspection is greatly reduced, and the real-time and continuity of monitoring data are difficult to guarantee. In addition, the traditional method has limited ability to capture the dynamic response of the support structure, and cannot realize the real-time linkage analysis of the support progress and the construction process, resulting in the risk of lag in project management.

[0003] With the development of computer vision technology, some studies have tried to achieve automatic monitoring of the support progress through technologies such as image recognition and 3D reconstruction. Existing solutions mostly use multi-camera arrays or lidar to build a 3D model of the slope, and calculate the deformation of the support structure through feature point matching algorithms. For example, the system based on a zoom vision displacement monitor realizes non-contact measurement through coded target recognition, but its deployment requires a large number of targets to be installed in advance and complex spatial coordinate system calibration to be completed. Although such a fully automated process can improve the monitoring efficiency, it has two significant defects: First, the high algorithm complexity leads to large consumption of computing resources. A single analysis requires the GPU cluster to perform parallel computing for dozens of minutes, and it is difficult to run in real time on the embedded devices at the construction site; Second, the system is too rigid. When the support plan is adjusted temporarily or sudden geological changes occur, it is necessary to reconfigure the monitoring network and train the model parameters, lacking adaptability to dynamic construction scenarios.

[0004] Therefore, there is an urgent need for a method for calculating the progress of slope support based on image processing, which has both flexibility and computing efficiency. Summary of the Invention

[0005] In order to solve the above technical problems, on the one hand, the present invention provides a method for calculating the progress of slope support based on image processing, including the following steps: S1: Obtain on-site pictures of slope support; the on-site pictures include the overall view of the slope support area and the time stamp when the pictures are collected; S2: According to the feature point identification criterion, obtain the feature points in the on-site picture through image recognition technology; S3: Taking the feature points in the on-site picture as the base points, combined with the slope support design drawing, correct the on-site picture into an on-site picture in the form of an orthographic view through homography transformation; S4: In the on-site picture in the form of an orthographic view, obtain the area of the slope support structure through a similarity algorithm; S5: Calculate the slope support progress based on the time stamp and the area of the slope support structure.

[0006] Preferably, the feature point identification criterion is to use the position of the drainage hole or the anchor rod as the feature point selected in the on-site picture.

[0007] Preferably, in S3, taking the feature points in the on-site picture as the base points, combined with the slope support design drawing, correct the on-site picture into an on-site picture in the form of an orthographic view through homography transformation, including the following steps: S3.1: According to the slope support design drawing, obtain the geometric coordinates of the feature points in the design coordinate system; S3.2: Establish a mapping relationship between the feature points identified in the on-site picture and the corresponding geometric coordinates in the design coordinate system; S3.3: Based on the mapping relationship, solve the image transformation matrix through the homography transformation algorithm; S3.4: Use the image transformation matrix to perform geometric correction on the on-site picture to generate an on-site picture in the form of an orthographic view with the same scale as the slope support design drawing.

[0008] Preferably, S4: In the on-site picture in the form of an orthographic view, obtain the area of the slope support structure through a similarity algorithm, including the following steps: S4.1: In the on-site picture in the form of an orthographic view, segment the slope support structure area through a similarity algorithm; S4.2: Generate a binary mask image for the segmented support structure area, mark the support area as the first pixel value, and mark the non-support area as the second pixel value; S4.3: Count the number of pixels corresponding to the support area in the binary mask image to obtain the pixel area; S4.4: According to the scale parameter of the orthographic view, convert the pixel area into the actual physical area.

[0009] Preferably, S4.1: In the on-site picture in the form of an orthographic view, segment the slope support structure area through a similarity algorithm, including: S4.1.1: Based on a preset seed point feature value threshold, initially select seed points in the on-site picture in the form of an orthographic view; S4.1.2: According to a preset similarity threshold, judge the difference between the feature values of the seed points and their respective adjacent pixels; S4.1.3: If the difference is less than the similarity threshold, add the adjacent pixel to the support area and use it as a new seed point; S4.1.4: Iteratively execute S4.1.2 to S4.1.3 until the entire on-site picture is traversed.

[0010] Preferably, according to the change in the area of the slope support structure at the time points corresponding to each time stamp, calculate the construction rate and the remaining construction time to achieve the calculation of the slope support progress.

[0011] On the other hand, this application also provides a slope support progress calculation system based on image processing, which executes the slope support progress calculation method based on image processing described in any one of the above. The calculation system includes: a feature point acquisition module, a picture conversion module, a slope support structure area calculation module, and a progress calculation module; The feature point acquisition module is used to obtain the feature points in the on-site picture through image recognition technology according to the feature point identification criterion; The picture conversion module is used to take the feature points in the on-site picture as the base points, combine with the slope support design drawing, and correct the on-site picture into an on-site picture in the form of an orthogonal view through homography transformation; The slope support structure area calculation module is used to obtain the area of the slope support structure through a similarity algorithm in the on-site picture in the form of an orthogonal view; The progress calculation module is used to calculate the slope support progress based on the time stamp and the area of the slope support structure.

[0012] On the other hand, this application also provides an electronic device, including a processor and a memory. The memory stores a computer program, and when the computer program is executed by the processor, it implements the slope support progress calculation method based on image processing described in any one of the above.

[0013] On the other hand, this application also provides a computer-readable storage medium, storing computer program instructions, and when the instructions are executed by a computing device, it implements the slope support progress calculation method based on image processing described in any one of the above.

[0014] The embodiments of the present invention have the following technical effects: A method for calculating the progress of slope support based on image processing provided by this application first involves obtaining on-site images of slope support with timestamps; then, according to the feature point identification criterion, through image recognition technology, obtaining the feature points in the on-site images; taking the feature points in the on-site images as the base points, combining with the slope support design drawing, and correcting the on-site images into on-site images in the form of an orthographic view through homography transformation; in the on-site images in the form of an orthographic view, obtaining the area of the slope support structure through a similarity algorithm; finally, calculating the slope support progress based on the timestamp and the area of the slope support structure. By using homography transformation to correct the on-site images with perspective distortion into an orthographic view, the influence of perspective differences on measurement is eliminated, and the two-dimensional images collected on-site are converted into measurement data with actual physical significance; the similarity algorithm identifies the segmented support structure area in the orthographic view. This application does not need to rely on complex algorithms, has high calculation efficiency, and does not need to reconfigure the monitoring network and train model parameters according to the adjustment of the support plan, jointly realizing efficient and flexible dynamic capture of the change of support progress and improving the response speed of project management. Description of the Drawings

[0015] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the description of the specific embodiments or the prior art. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0016] Figure 1 It is a flowchart of the steps of the method for calculating the progress of slope support provided by the embodiment of the present invention; Figure 2 It is a schematic diagram of the feature points in the on-site image before correction provided by the embodiment of the present invention; Figure 3 It is a schematic diagram of the feature points in the on-site image in the form of an orthographic view after correction provided by the embodiment of the present invention; Figure 4 It is a schematic diagram of the marked seed points for framing the area of concrete provided by the embodiment of the present invention; Figure 5 It is a schematic diagram of the area of the slope support structure segmented by the similarity algorithm provided by the embodiment of the present invention. Detailed Embodiments

[0017] To make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be described clearly and completely below. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope protected by the present invention.

[0018] The present invention provides a method for calculating the progress of slope support based on image processing, aiming to solve the problems of low efficiency and poor flexibility of traditional monitoring means. This method converts on-site pictures into orthographic views that can be quantitatively analyzed through image recognition and geometric correction technologies, combines similarity algorithms to calculate the support area, and finally realizes the dynamic monitoring of the construction progress.

[0019] Figure 1 It is the flowchart of the steps of the method for calculating the progress of slope support provided by the embodiments of the present invention. Refer to Figure 1 , and specifically includes the following steps: S1: Obtain on-site pictures of slope support; the on-site pictures include the overall view of the slope support area and the time stamp when the pictures are collected; S2: According to the feature point identification criterion, obtain the feature points in the on-site pictures through image recognition technology; In some embodiments, the feature point identification criterion is to use the positions of drainage holes or anchor bolts as the feature points selected in the on-site pictures.

[0020] Drainage holes and anchor bolts are typical predefined features in slope support structures. Their positions are usually clearly marked in the design drawings, and the actual positions after construction are highly consistent with the design drawings. Selecting such feature points can ensure the matching accuracy and reduce the adaptability requirements of the algorithm for complex scenarios. The geometric positions of the predefined feature points are stable, reducing the matching error. Drainage holes and anchor bolts are not easily deformed or moved after construction and are suitable as reference points for long-term monitoring, improving the robustness of progress calculation. Preferably, there should be at least 4 drainage holes or anchor bolts in the photo to form the feature points of the transformation matrix.

[0021] Exemplarily, during implementation, the system first identifies the drainage holes or anchor bolts in the on-site pictures. For drainage holes, they can be located through circular detection algorithms (such as Hough transform); for anchor bolts, their contours are extracted by combining edge detection and morphological processing.

[0022] S3: Using the feature points in the on-site pictures as the base points, combined with the slope support design drawings, correct the on-site pictures into on-site pictures in the form of orthographic views through homography transformation; Figure 2 It is a schematic diagram of the feature points in the on-site pictures before correction provided by the embodiments of the present invention, where the position of the anchor bolt is used as the feature point selected in the on-site pictures; Figure 3 isFigure 2 Schematic diagram of feature points in the on-site picture in the form of a corrected orthographic view; In some embodiments, in S3, taking the feature points in the on-site picture as the base points, and combining with the slope support design drawing, the on-site picture is corrected into an on-site picture in the form of an orthographic view through a homography transformation, including the following steps: S3.1: According to the slope support design drawing, obtain the geometric coordinates of the feature points in the design coordinate system; S3.2: Establish a mapping relationship between the feature points identified in the on-site picture and the corresponding geometric coordinates in the design coordinate system; S3.3: Based on the mapping relationship, solve the image transformation matrix through the homography transformation algorithm; S3.4: Use the image transformation matrix to perform geometric correction on the on-site picture to generate an on-site picture in the form of an orthographic view with the same scale as the slope support design drawing. The positions and spacings of the drainage holes or anchor rods in the orthographic view are consistent with the design drawings.

[0023] Geometric correction is performed on the on-site picture through homography transformation to generate an orthographic view with the same scale as the design drawing, thereby eliminating geometric distortion caused by factors such as camera perspective and perspective distortion during image acquisition, and providing a standardized and quantifiable image basis for subsequent support area calculation and progress analysis. The key feature of the orthographic view is that its projection method can maintain a fixed proportional relationship between the object size and the actual physical size, enabling the measurement results in the image to be directly mapped to the actual engineering scenario.

[0024] S4: In the on-site picture in the form of an orthographic view, obtain the area of the slope support structure through a similarity algorithm; In some embodiments, S4: In the on-site picture in the form of an orthographic view, obtain the area of the slope support structure through a similarity algorithm, including the following steps: S4.1: In the on-site picture in the form of an orthographic view, segment the slope support structure area through a similarity algorithm; In some embodiments, S4.1: In the on-site picture in the form of an orthographic view, segment the slope support structure area through a similarity algorithm, including: S4.1.1: Based on a preset seed point eigenvalue threshold, initially select seed points in the on-site picture in the form of an orthographic view, as Figure 4 shown; Exemplarily, the preset seed points represent suspected support structure concrete areas. In the on-site picture in the form of an orthographic view, randomly select points whose color or texture is similar to that of the concrete to initially select the seed points. The eigenvalue threshold of the preset seed points is the color or texture of the seed points.

[0025] S4.1.2: According to the preset similarity threshold, judge the difference in the eigenvalues of the seed points and their adjacent pixels; Exemplarily, before performing this step, it is necessary to determine whether the distance between the seed point and its respective adjacent pixels meets a preset distance threshold. If it is greater than the distance threshold, then S4.1.2 does not need to be executed.

[0026] S4.1.3: If the difference is less than the similarity threshold, then add the adjacent pixel to the support area and use it as a new seed point, that is, a growth point; If the difference is greater than or equal to the similarity threshold, then do not add the adjacent pixel to the support area.

[0027] S4.1.4: Iteratively execute S4.1.2 to S4.1.3 until the entire on-site picture is traversed.

[0028] Exemplarily, when segmenting the slope support structure area through a similarity algorithm, the support area can be automatically extended and boxed based on color, texture, or edge features. Multiple similarity features (such as texture, edge) are adapted to different support materials.

[0029] S4.2: Generate a binary mask image for the segmented support structure area, and mark the support area with a first pixel value and the non-support area with a second pixel value; Exemplarily, the generation of the binary mask image includes performing a morphological closing operation on the segmented area to eliminate noise or holes.

[0030] Exemplarily, count the number of white pixels in the mask image and convert it to a physical area. The key code is as follows: area_pixels = np.sum(mask == 255); To distinguish black and white in the figure, white is the pixel area to be calculated, and black is the pixel area that does not need to be calculated. Use the above code to count the number of white pixels in the mask image and convert it to a physical area. S4.3: Count the number of pixels corresponding to the support area in the binary mask image to obtain the pixel area; Exemplarily, the pixel area is counted by traversing the binary mask image and accumulating the number of pixels marked as the support area. S4.4: According to the scale parameter of the orthographic view, convert the pixel area to an actual physical area.

[0031] Exemplarily, the scale parameter is determined by the ratio of the actual distance between feature points in the orthographic view to the pixel pitch. Select the actual physical distances between at least two feature points from the slope support design drawing. The actual physical distances between the feature points are obtained based on the annotation data in the design drawing, including the arrangement distances of drainage holes or anchor rods. In the on-site picture in the form of an orthographic view, measure the pixel pitch between the two feature points. The measurement of the pixel pitch is achieved by calculating the Euclidean distance of the coordinate differences of the feature points in the orthographic view. According to the ratio of the actual physical distance to the pixel pitch, calculate the actual physical length corresponding to a unit pixel. If there are multiple pairs of feature points, calculate the average value through the ratios of multiple groups of actual physical distances to pixel pitches as the final scale parameter. The X-axis and Y-axis directions of the scale parameter are calculated independently to ensure the isotropy of the scale of the corrected image.

[0032] Exemplarily, for the seed points selected by the system, the region growing is implemented using the region_growing_with_threshold function to determine whether they belong to similar regions and mark them as white. The key code is as follows: if abs(int(img[y, x]) - int(seed_value))<= threshold: mask[y, x] = 255 region_pixels.append((x, y)) overlay[y, x] = [255, 255, 255]; The meaning of this section of the growth code is that first, a row of seed points has been marked, as Figure 4 shown. Then, compare adjacent pixels. If the difference between an adjacent pixel and the seed point is less than the threshold, then mark the mask of the adjacent pixel as 255, that is, white. Taking the pixels from the previous step as the basis, continue to expand outward until the entire picture is traversed.

[0033] Exemplarily, region segmentation: In the orthographic view, select the seed points (such as the concrete surface) within the support area. The system automatically expands the area according to a preset similarity threshold (such as the color difference is less than 10%). Generate a mask: Mark the segmented area as white (pixel value 255) and the non-support area as black (pixel value 0) to form a binary image. Pixel statistics: Traverse the mask image and accumulate the number of white pixels to obtain the pixel area of the support area. Unit conversion: According to the scale parameter (such as 1 pixel = 0.01 m²), convert the pixel area into the actual physical area. Eliminate holes or noise in the image through morphological closing operations to improve the accuracy of area calculation.

[0034] S5: Calculate the slope support progress based on the timestamp and the area of the slope support structure.

[0035] In some embodiments, according to the change in the area of the slope support structure at the time points corresponding to each timestamp, the construction rate and the remaining construction time are calculated to achieve the calculation of the slope support progress.

[0036] Exemplarily, the construction rate is calculated according to the time interval and the area difference: 。

[0037] Exemplarily, according to the photos at different time points, the change in the support area is analyzed, the construction rate is calculated, and the completion time is predicted. Assuming that the construction rate remains constant with the construction rates of the last two time intervals, the remaining time is calculated according to the total target support area: 。

[0038] Therefore, the slope support progress calculation method provided by the present application does not require professional equipment or complex deployment, has high efficiency, can shorten the monitoring period; high-precision measurement can be achieved by using ordinary imaging equipment, reducing the hardware investment and realizing low-cost progress monitoring; continuous monitoring is supported, the change in the support progress is dynamically captured, the response speed of project management is improved, and the real-time performance is high.

[0039] On the other hand, the present application also provides a slope support progress calculation system based on image processing, which executes the slope support progress calculation method based on image processing described in any one of the above. The calculation system includes: a feature point acquisition module, a picture conversion module, a slope support structure area calculation module, and a progress calculation module; The feature point acquisition module is used to obtain the feature points in the on-site picture through image recognition technology according to the feature point identification criterion; The picture conversion module is used to correct the on-site picture into an on-site picture in the form of an orthographic view through homography transformation with the feature points in the on-site picture as the base points and in combination with the slope support design drawing; The slope support structure area calculation module is used to obtain the area of the slope support structure through a similarity algorithm in the on-site picture in the form of an orthographic view; The progress calculation module is used to calculate the slope support progress based on the timestamp and the area of the slope support structure.

[0040] On the other hand, the present application also provides an electronic device, including a processor and a memory. The memory stores a computer program, and when the computer program is executed by the processor, the slope support progress calculation method based on image processing described in any one of the above is implemented.

[0041] On the other hand, the present application also provides a computer-readable storage medium storing computer program instructions, which, when executed by a computing device, implement the method for calculating the progress of slope support based on image processing described in any one of the above.

[0042] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the technical solutions of the embodiments of the present invention.

Claims

1. A slope support progress calculation method based on image processing, characterized in that: The steps include: S1: Obtaining a slope support site picture; the site picture includes a full view of the slope support area and a timestamp when the picture was collected; S2: Acquire feature points in the scene picture by image recognition technology according to feature point identification criteria; S3: Taking the characteristic points in the site picture as base points and combining with the slope support design drawing, the site picture is corrected into a site picture in the form of an orthogonal view through homography transformation; S4: in the on-site picture in the form of an orthogonal view, obtaining the area of ​​the slope support structure by a similarity algorithm; S5: Calculating slope support progress based on the timestamp and the area of ​​the slope support structure.

2. The slope support progress calculation method based on image processing according to claim 1 is characterized in that: The characteristic point identification criterion is to use the drainage hole or the anchor rod position as the characteristic point selected from the site picture.

3. The slope support progress calculation method based on image processing according to claim 1 is characterized in that: In S3, the feature points in the site picture are used as base points, and the site picture is corrected into an orthogonal view through homography transformation in combination with the slope support design drawing, including the following steps: S3.1: According to the slope support design drawing, the geometric coordinates of the feature points in the design coordinate system are obtained; S3.2: A mapping relationship is established between the feature points identified in the site picture and the corresponding geometric coordinates in the design coordinate system; S3.3: Based on the mapping relationship, the image transformation matrix is ​​solved by the homography transformation algorithm; S3.4: The site picture is geometrically corrected using the image transformation matrix to generate a site picture in the form of an orthogonal view that is consistent with the proportions of the slope support design drawing.

4. The slope support progress calculation method based on image processing according to claim 1 is characterized in that: S4: In the on-site picture in the form of an orthogonal view, the area of ​​the slope support structure is obtained by a similarity algorithm, comprising the following steps: S4.1: In the site picture in the orthogonal view form, the slope support structure area is segmented by a similarity algorithm; S4.2: Generate a binary mask image for the segmented support structure area, and mark the support area as a first pixel value and the non-support area as a second pixel value; S4.3: Count the number of pixels corresponding to the support area in the binary mask image to obtain the pixel area; S4.4: Convert the pixel area into an actual physical area according to the scale parameter of the orthogonal view.

5. The slope support progress calculation method based on image processing according to claim 4 is characterized in that: S4.1: In the orthogonal view of the site image, the slope support structure area is segmented by a similarity algorithm, including: S4.1.1: Initially selecting a seed point in the scene image in the orthogonal view form based on a preset seed point feature value threshold; S4.1.2: Determine the difference between the feature values ​​of the seed point and each adjacent pixel according to a preset similarity threshold; S4.1.3: If the difference is less than the similarity threshold, adding the adjacent pixel to the support area and serving as a new seed point; S4.1.4: Iterate S4.1.2 to S4.1.3 until the entire scene picture is traversed.

6. The slope support progress calculation method based on image processing according to claim 1 is characterized in that: According to the change in the area of ​​the slope support structure at the time point corresponding to each time stamp, the construction rate and the remaining construction time are calculated to realize the calculation of the slope support progress.

7. A slope support progress calculation system based on image processing, characterized in that: Execute a slope support progress calculation method based on image processing as described in any one of claims 1 to 6, the calculation system comprising: a feature point acquisition module, an image conversion module, a slope support structure area calculation module, and a progress calculation module; A feature point acquisition module is used to acquire feature points in the scene picture through image recognition technology according to feature point identification criteria; A picture conversion module is used to correct the scene picture into a scene picture in the form of an orthogonal view by homography transformation based on the characteristic points in the scene picture as base points and in combination with the slope support design drawing; A slope support structure area calculation module is used to obtain the area of ​​the slope support structure in the on-site picture in the form of an orthogonal view by a similarity algorithm; A progress calculation module is used to calculate the slope support progress based on the timestamp and the area of ​​the slope support structure.

8. An electronic device, characterized in that: The method comprises a processor and a memory, wherein the memory stores a computer program. When the computer program is executed by the processor, a slope support progress calculation method based on image processing as described in any one of claims 1 to 6 is implemented.

9. A computer-readable storage medium, characterized in that: Computer program instructions are stored, and when the instructions are executed by a computing device, a slope support progress calculation method based on image processing as described in any one of claims 1 to 6 is implemented.

Images