A method for constructing a digital three-dimensional model of a coal mine filling mining working face

Abstract

The present invention discloses a method for constructing a digital three-dimensional model of a coal mine filling mining working face, and relates to the field of image processing technology. The method comprises: obtaining a first working face image and a second working face image of coal mine filling mining; determining the filling expression degree of each working face area in the first working face image and the second working face image respectively according to the gray value of each pixel point in the first working face image and the second working face image; determining the boundary line matching degree between the first dividing line of the first working face image and the second dividing line of the second working face image according to the filling expression degree of each working face area; correcting the initial matching degree of the matching point pair between the first working face image and the second working face image according to the matching degree of each dividing line to obtain the final matching degree; using the final matching degree of each matching point pair, splicing the first working face image and the second working face image to form a digital three-dimensional model of the coal mine filling mining working face.

Description

Technical Field

[0001] The invention relates to the technical field of image processing, and in particular to a method for constructing a digital three-dimensional model of a coal mine filling mining working face. Background Art

[0002] Backfill mining is a green mining technology that can effectively control risks. However, the implementation of backfill mining involves complex geological conditions, backfill material characteristics, construction progress, safety monitoring and other factors. Therefore, it is very important to conduct digital 3D modeling of the backfill mining working face in coal mines.

[0003] In the existing methods, based on image processing and computer vision technology, the coal mine filling mining working face is simulated and constructed through technical means such as image acquisition and processing, three-dimensional reconstruction and dynamic updating.

[0004] However, in this method, due to the inconsistent angles of image acquisition, terrain complexity, texture repetitiveness or large changes in viewing angles, the overlapping parts of the images are difficult to accurately align, resulting in obvious seams, distortion or misalignment during the stitching process, resulting in poor accuracy of the final three-dimensional model. Summary of the invention

[0005] The embodiment of the present invention provides a method for constructing a digital three-dimensional model of a coal mine filling mining working face, which can improve the accuracy of the ultimately constructed three-dimensional model.

[0006] A first aspect of an embodiment of the present invention provides a method for constructing a digital three-dimensional model of a coal mine backfilling working face, comprising:

[0007] Acquire a first working face image and a second working face image of coal mine backfill mining, wherein the first working face image and the second working face image respectively include a plurality of working face areas and a plurality of edge lines, and the first working face image and the second working face image are images of two adjacent areas in the coal mine backfill mining working face;

[0008] Determine the filling expression degree of each working surface area in the first working surface image and the second working surface image according to the gray value of each pixel point in the first working surface image and the second working surface image, respectively, and the filling expression degree is used to characterize the possibility that the working surface area belongs to the filling body area;

[0009] According to the filling expression degree of each working surface area, determining the boundary line matching degree between the first boundary line of the first working surface image and the second boundary line of the second working surface image, the boundary line being the edge line between the filling body area and other areas among the multiple edge lines of the working surface image;

[0010] According to the matching degree of each dividing line, the initial matching degree of the matching point pair between the first working surface image and the second working surface image is corrected to obtain the final matching degree of the matching point pair between the first working surface image and the second working surface image;

[0011] The final matching degree of each matching point pair is used to splice the first working face image with the second working face image to form a digital three-dimensional model of the coal mine filling mining working face.

[0012] In the method for constructing a digital three-dimensional model of a coal mine backfilling working face provided by an embodiment of the present invention, the working face image is first segmented to determine the possibility that each working face area belongs to the backfill area after segmentation, that is, the filling performance degree. The working face can be effectively divided into different areas, thereby reducing the interference of repeated textures in the working face image and improving the matching accuracy of the matching point pairs. Then, according to the filling performance degree of each working face area, the matching degree of the boundary line between the first dividing line of the first working face image and the second dividing line of the second working face image is accurately identified. By analyzing the positional relationship between the dividing lines of different areas, the relative position and posture between different working face images can be further determined, and the matching accuracy of the matching point pairs can be improved. In this way, according to the matching degree of the dividing line, the initial matching degree of the matching point pairs between the first working face image and the second working face image is corrected, which can improve the matching accuracy of the matching point pairs. Thereby, the precise alignment of the overlapping parts between the first working face image and the second working face image can be achieved, thereby improving the accuracy of the finally constructed three-dimensional model. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0014] Figure 1 A schematic flow chart of a first method for constructing a digital three-dimensional model of a coal mine backfilling working face provided by an embodiment of the present invention;

[0015] Figure 2 A schematic flow chart of a second method for constructing a digital three-dimensional model of a coal mine backfilling working face provided by an embodiment of the present invention;

[0016] Figure 3 A schematic flow chart of a third method for constructing a digital three-dimensional model of a coal mine filling mining working face provided by an embodiment of the present invention;

[0017] Figure 4A schematic flow chart of a fourth method for constructing a digital three-dimensional model of a coal mine filling mining working face provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0018] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the method for constructing a digital three-dimensional model of a coal mine filling mining working face proposed by the present invention, its specific implementation method, structure, characteristics and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0019] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0020] It should be noted that the acquisition, storage, use, and processing of data in the technical solution of the present invention are in compliance with the relevant provisions of laws and regulations.

[0021] It should be noted that in the embodiments of the present invention, certain software, components, models and other existing solutions in the industry may be mentioned, which should be regarded as exemplary. Their purpose is only to illustrate the feasibility of implementing the technical solution of the present invention, but it does not mean that the applicant has or will necessarily use the solution.

[0022] In the existing methods, based on image processing and computer vision technology, the coal mine filling mining working face is simulated and constructed through technical means such as image acquisition and processing, three-dimensional reconstruction and dynamic update. However, this method easily leads to poor accuracy of the final three-dimensional model.

[0023] The object of the present invention is to provide a method for constructing a digital three-dimensional model of a coal mine filling mining working face. In the method for constructing a digital three-dimensional model of a coal mine filling mining working face provided by the embodiment of the present invention, the working face image is first segmented to determine the possibility that each working face area belongs to the filling body area after segmentation, that is, the filling performance degree. The working face can be effectively divided into different areas, thereby reducing the interference of repeated textures in the working face image and improving the matching accuracy of the matching point pairs. Then, according to the filling performance degree of each working face area, the boundary matching degree between the first dividing line of the first working face image and the second dividing line of the second working face image is accurately identified. By analyzing the positional relationship between the dividing lines of different areas, the relative position and posture between different working face images can be further determined, and the matching accuracy of the matching point pairs can be improved. In this way, according to the boundary matching degree, the initial matching degree of the matching point pairs between the first working face image and the second working face image is corrected, which can improve the matching accuracy of the matching point pairs. Thereby, the precise alignment of the overlapping part between the first working face image and the second working face image can be achieved, thereby improving the accuracy of the finally constructed three-dimensional model.

[0024] The following describes a specific embodiment of a method for constructing a digital three-dimensional model of a coal mine filling mining working face provided by an embodiment of the present invention.

[0025] Figure 1 A flow chart of a method for constructing a digital three-dimensional model of a coal mine filling mining face is provided. The method for constructing a digital three-dimensional model of a coal mine filling mining face can be applied to a server. The method for constructing a digital three-dimensional model of a coal mine filling mining face can include the following S101 to S105.

[0026] S101, obtaining a first working face image and a second working face image of coal mine filling mining, wherein the first working face image and the second working face image respectively include a plurality of working face areas and a plurality of edge lines, and the first working face image and the second working face image are images of two adjacent areas in the coal mine filling mining working face.

[0027] In this embodiment, the first working surface image and the second working surface image are images of two adjacent areas in the working surface of coal mine filling mining, and there is a certain overlapping area between the first working surface image and the second working surface image, so that enough feature points can be found when the first working surface image and the second working surface image are subsequently spliced.

[0028] As an example, according to the size and shape of the coal mine backfill mining working face, multiple camera fixing points are planned to ensure that the camera field of view at each camera fixing point corresponds to different areas of the working face.

[0029] Then, for each camera at the fixed point, the server sends an image acquisition command to make the camera rotate from left to right in 20° increments starting from the preset fixed position. After each adjustment, an image of the work surface is taken to ensure that there is a certain overlap between the images of the work surface taken before and after, until the area corresponding to the camera is traversed.

[0030] Then, the server uniformly names and classifies the images according to the shooting point, shooting angle, shooting time, etc., and uploads the collected work surface images to the cloud platform for preprocessing. Specifically, it can include denoising (using filtering algorithms to remove environmental noise), brightness and contrast adjustment (enhancing image details in low-light environments), distortion correction (correcting geometric deformation caused by the lens according to camera calibration parameters), image cropping and unified resolution (ensuring consistent image size for subsequent stitching), and color balancing (eliminating color cast caused by different light sources) and other operations to optimize image quality.

[0031] Finally, the server selects two working surface images with a certain overlapped area from each working surface image as the first working surface image and the second working surface image. At the same time, the Otsu threshold segmentation method is used to divide the first working surface image and the second working surface image into multiple working surface areas; edge detection is performed on the first working surface image and the second working surface image to obtain multiple edge lines in the first working surface image and the second working surface image.

[0032] S102, determining the filling expression degree of each working surface area in the first working surface image and the second working surface image according to the gray value of each pixel point in the first working surface image and the second working surface image, respectively, wherein the filling expression degree is used to characterize the possibility that the working surface area belongs to the filling body area.

[0033] In this embodiment, the working face image includes multiple regions, such as a filling body region, a coal seam region, and a rock layer region, etc. The texture and illumination characteristics of different regions are different.

[0034] The filling performance degree reflects the possibility that the working surface area belongs to the filling body area. The greater the filling performance degree, that is, the brighter the working surface area is, the greater the possibility that the working surface area belongs to the filling body area.

[0035] As an example, the server uses an image processing tool to extract the grayscale value of each pixel in the first working surface image and the second working surface image, and then sets a threshold value based on the distribution range of the grayscale value to distinguish the filling body area from other areas.

[0036] Finally, for each working surface area, the matching degree between the grayscale value of its internal pixel and the threshold is calculated. The higher the matching degree, the greater the possibility that the working surface area belongs to the filling body area, that is, the higher the filling performance.

[0037] S103, determining the degree of boundary line matching between a first boundary line of the first working surface image and a second boundary line of the second working surface image according to the filling expression degree of each working surface area, wherein the boundary line is an edge line between the filling body area and other areas among multiple edge lines of the working surface image.

[0038] In this embodiment, the boundary line is used to characterize the edge line between the filling body area and other areas in the working surface image. The first boundary line is the boundary line in the first working surface image, and the second boundary line is the boundary line in the second working surface image.

[0039] The boundary line matching degree is used to represent the matching degree between the first boundary line in the first working surface image and the second boundary line in the second working surface image.

[0040] As an example, the server determines the working surface area with the highest filling degree in the first working surface image as the filling body area in the first working surface image, and determines the working surface area with the highest filling degree in the second working surface image as the filling body area in the second working surface image based on the filling degree of each working surface area in the first working surface image and the second working surface image.

[0041] Then, according to the filling body area in the first working surface image, each edge in the first working surface image is screened to obtain at least one first dividing line; according to the filling body area in the second working surface image, each edge in the second working surface image is screened to obtain at least one second dividing line.

[0042] Finally, the first dividing line of the first working surface image is compared with the corresponding second dividing line of the second working surface image one by one. The similarity of the shape, position, length and other features between the two dividing lines is calculated, and the weighted sum is performed to determine the degree of matching between them.

[0043] S104, correcting the initial matching degree of the matching point pairs between the first working surface image and the second working surface image according to the matching degree of each boundary line, to obtain the final matching degree of the matching point pairs between the first working surface image and the second working surface image.

[0044] In this embodiment, the matching point pair is used to provide a basis for image stitching, and includes a feature point belonging to the overlapping area in the first working surface image and a corresponding feature point belonging to the overlapping area in the second working surface image.

[0045] As an example, the server first uses a feature point matching algorithm in an image matching algorithm to calculate an initial matching degree of matching point pairs between the first working surface image and the second working surface image.

[0046] Then, according to the matching degree of each dividing line, the initial matching degree of each matching point pair between the first working surface image and the second working surface image is corrected. Specifically, the higher the average value of the dividing line matching degree, the more similar the positional relationship of the dividing line between the first working surface image and the second working surface image is, and therefore the matching degree of the corresponding matching point pair should be greater.

[0047] Finally, the matching degree of the matching point pairs between the corrected first working surface image and the second working surface image is taken as the final matching degree for subsequent image stitching and three-dimensional model construction.

[0048] S105, using the final matching degree of each matching point pair, splicing the first working surface image and the second working surface image to form a digital three-dimensional model of the coal mine filling mining working surface.

[0049] In this embodiment, the server uses an image stitching algorithm to stitch the first working surface image and the second working surface image according to the final matching degree of each matching point pair. During the stitching process, it is necessary to ensure smooth transition and seamless connection between the images.

[0050] After all the first working surface images and the second working surface images are spliced ​​in the above manner to obtain a complete coal mine filling mining working surface image, a three-dimensional model is constructed using a three-dimensional modeling software or algorithm. Specifically, the three-dimensional shape and structure of the coal mine filling mining working surface area are restored based on the grayscale value, edge line and other information of the complete coal mine filling mining working surface image. At the same time, the three-dimensional model can be made more realistic and true by adding texture mapping, lighting processing and other means.

[0051] In the method for constructing a digital three-dimensional model of a coal mine backfilling working face provided in this embodiment, the working face image is first segmented to determine the possibility that each working face area belongs to the backfill area after segmentation, that is, the filling performance degree. The working face can be effectively divided into different areas, thereby reducing the interference of repeated textures in the working face image and improving the matching accuracy of the matching point pairs. Then, according to the filling performance degree of each working face area, the matching degree of the boundary line between the first dividing line of the first working face image and the second dividing line of the second working face image is accurately identified. By analyzing the positional relationship between the dividing lines of different areas, the relative position and posture between different working face images can be further determined, and the matching accuracy of the matching point pairs can be improved. In this way, according to the matching degree of the dividing line, the initial matching degree of the matching point pairs between the first working face image and the second working face image is corrected, which can improve the matching accuracy of the matching point pairs. Thereby, the precise alignment of the overlapping parts between the first working face image and the second working face image can be achieved, thereby improving the accuracy of the finally constructed three-dimensional model.

[0052] As an optional embodiment, Figure 2 As shown, S102 may specifically include the following S201 to S202:

[0053] For each working surface area in the first working surface image and the second working surface image, the following steps are performed respectively:

[0054] S201, determining a color contrast of a target working surface area according to a grayscale value of each pixel in the target working surface area, where the target working surface area is any one of the various working surface areas, and the color contrast is used to characterize the color vividness of the target working surface area;

[0055] S202, determining a filling expression degree of the target working surface area according to a color contrast of the target working surface area and a grayscale value of each pixel in the target working surface area.

[0056] In this embodiment, the target working surface area is a specific area of ​​interest in the working surface image, which may be any one of the various working surface areas.

[0057] Grayscale value is a numerical value representing the brightness of each pixel in an image, and the value range is usually from 0 (black) to 255 (white).

[0058] Color contrast is a measure of the color differences in the work surface area. Higher color contrast means that there is a significant brightness change in the area, that is, the work surface area is visually more vivid or bright.

[0059] As an example, the server determines the color contrast of the target work surface area by calculating the difference or distribution of the grayscale values ​​of all pixels in the target work surface area. Specifically, the variance or standard deviation of the grayscale histogram of the target work surface area can be calculated, or other statistical methods can be used to quantify the distribution and variation of the grayscale values ​​in the target work surface area.

[0060] Then, the filling performance degree of the target working surface area is calculated according to the color contrast of the target working surface area and the grayscale value of each pixel in the target working surface area. Specifically, the greater the color contrast, the greater the corresponding filling performance degree; the more uneven the grayscale value distribution of each pixel in the target working surface area, the greater the corresponding filling performance degree.

[0061] Through this embodiment, the color contrast of the target working surface area is determined according to the grayscale value of each pixel in the target working surface area. Then, the filling performance degree of the target working surface area is determined according to the color contrast of the target working surface area and the grayscale value of each pixel in the target working surface area. In this way, by accurately calculating the filling performance degree of the target working surface area, it is possible to accurately determine whether the target working surface area belongs to the filling body area.

[0062] As an optional embodiment, S201 may specifically include:

[0063] The grayscale value of each pixel in the target working surface area is counted to obtain the average grayscale value, the maximum grayscale value and the minimum grayscale value in the target working surface area;

[0064] The difference between the average gray value and the minimum gray value is divided by the difference between the maximum gray value and the minimum gray value, and the absolute value is taken to obtain a first calculation result;

[0065] The first calculation result is multiplied by the average gray value to obtain the color contrast of the target working surface area.

[0066] In this embodiment, the filling area of ​​the coal mine appears as a relatively obvious bright area in the coal mine working face due to the material of the material, so the filling area and other areas can be distinguished based on the difference in color brightness.

[0067] As an example, the color contrast ratio may be specifically determined by the following formula 1:

[0068] Formula 1

[0069] In the formula, The color contrast used to characterize the i-th working surface area, It is used to represent the average gray value of all pixels in the i-th working surface area. It is used to represent the minimum gray value of the pixel point in the i-th working surface area. It is used to represent the maximum gray value of the pixel points in the i-th working surface area.

[0070] Among them, the average gray value of all pixels in the i-th working surface area The larger it is, the brighter the overall color in the i-th working surface area is, and the greater the color contrast in the i-th working surface area is; the greater the difference between the maximum grayscale value and the minimum grayscale value in the i-th working surface area is, the greater the color fluctuation in the i-th working surface area is, which means that the possibility that the overall color in the i-th working surface area is brighter is lower, that is, the color contrast in the i-th working surface area is smaller.

[0071] Color contrast of the i-th work surface area Used to characterize the overall color brightness of the i-th working surface area, The larger the value is, the brighter the overall color of the i-th working surface area is, that is, the greater the possibility that the i-th working surface area belongs to the filling body area.

[0072] In this embodiment, the color contrast of the target working surface area is accurately calculated according to the gray value of each pixel in the target working surface area. Therefore, the possibility that the target working surface area belongs to the filling body area can be determined according to the color contrast of the target working surface area, which can help to accurately judge whether the target working surface area belongs to the filling body area.

[0073] As an optional embodiment, S202 may specifically include:

[0074] Based on the grayscale value of each pixel in the target working surface area, the grayscale value difference between adjacent pixels in the target working surface area is calculated;

[0075] Calculate the average value of each gray value difference to obtain a second calculation result;

[0076] The second calculation result is multiplied by the color contrast of the target working surface area to obtain the filling expression degree of the target working surface area.

[0077] In this embodiment, the filling performance degree can be specifically determined by the following formula 2:

[0078] Formula 2

[0079] In the formula, Used to characterize the filling performance of the i-th working surface area, The color contrast used to characterize the i-th working surface area, Used to represent the gray value of the nth pixel in the i-th working surface area, It is used to represent the gray value of the n+1th pixel in the i-th working surface area, and N is used to represent the total number of pixels in the i-th working surface area.

[0080] in, It is used to characterize the second calculation result obtained by calculating the average value of the gray value difference between each adjacent pixel point, which is used to reflect the chaos of the gray value in the i-th working surface area. Specifically, in addition to the brighter color of the filling body area in the coal mine, there will also be brighter areas on the metal surface of the mining machine, but the filling body area presents a bright color with a relatively chaotic color, while the metal surface of the mining machine presents a bright color with a relatively consistent color. Therefore, the more chaotic the gray value in the i-th working surface area, the greater the possibility that the i-th working surface area belongs to the filling body area, and the greater the corresponding filling performance; the brighter the overall color of the i-th working surface area, the greater the possibility that the i-th working surface area belongs to the filling body area, and the greater the corresponding filling performance.

[0081] The filling performance of the i-th working surface area The larger the value is, the more likely it is that the i-th working surface area belongs to the filling body area. In this way, by calculating the filling performance of each working surface area, the real filling body area can be determined.

[0082] Through this embodiment, the filling performance degree of the target working surface area is calculated based on the color contrast of the target working surface area and the grayscale value of each pixel in the target working surface area. In this way, a comprehensive evaluation is performed from two aspects: color brightness and grayscale value confusion, so that it is possible to accurately determine whether the target working surface area belongs to the filling body area, thereby improving the accuracy of the filling body area judgment.

[0083] As an optional embodiment, Figure 3 As shown, S103 may specifically include the following S301 to S304:

[0084] S301, based on the filling expression degree of each working surface area, each edge line in the first working surface image and the second working surface image is screened to obtain a first dividing line in the first working surface image and a second dividing line in the second working surface image;

[0085] S302, determining boundary line similarity between the first boundary line and the second boundary line according to the distance between the first boundary line and the second boundary line;

[0086] S303, determining the direction consistency of the first dividing line and the second dividing line according to the gradient angle difference between each adjacent pixel point in the first dividing line and the gradient angle difference between each adjacent pixel point in the second dividing line;

[0087] S304: Calculate the boundary line matching degree between the first boundary line and the second boundary line based on the boundary line similarity and direction consistency.

[0088] In this embodiment, the boundary line similarity is used to characterize the possibility that the first boundary line and the second boundary line determined according to the positional relationship belong to the boundary line of the same filling body region.

[0089] Directional consistency is used to characterize the degree to which the changes of the first dividing line and the second dividing line are at a consistent angle, that is, the degree to which the change trends are consistent.

[0090] As an example, the server determines the working surface area with the highest filling degree in the first working surface image as the filling body area in the first working surface image, and determines the working surface area with the highest filling degree in the second working surface image as the filling body area in the second working surface image based on the filling degree of each working surface area in the first working surface image and the second working surface image.

[0091] Then, according to the filling body area in the first working surface image, each edge in the first working surface image is screened to obtain at least one first dividing line; according to the filling body area in the second working surface image, each edge in the second working surface image is screened to obtain at least one second dividing line.

[0092] Then, for each point on the first dividing line, find the nearest point on the second dividing line and calculate the distance between them. Specifically, this can be achieved by traversal or nearest neighbor search algorithm. And calculate the average or median of all point distances as a measure of the dividing line similarity between the first dividing line and the second dividing line. Among them, the smaller the distance, the higher the dividing line similarity.

[0093] Then, for each pixel point on the first dividing line and the second dividing line, obtain their gradient angles along the edge tangent direction, and subtract the gradient angles of two adjacent pixel points to obtain the gradient angle difference between adjacent pixel points. Then, according to the gradient angle difference between adjacent pixel points, the direction consistency of the first dividing line and the second dividing line is determined specifically by the following formula 3:

[0094] Formula 3

[0095] In the formula, Used to characterize the direction consistency between the j-th first dividing line of the a-th first working surface image and the k-th second dividing line of the b-th second working surface image. It is used to characterize the gradient angle difference between the mth group of adjacent pixel points in the jth first dividing line of the ath first working surface image, It is used to represent the gradient angle difference between the mth group of adjacent pixels in the kth second dividing line of the bth second working surface image. M is used to represent the total number of groups of adjacent pixels on the shorter side of the first dividing line and the second dividing line, and exp is used to represent the exponential function operation.

[0096] in, It is used to characterize the difference in gradient angle between the jth first dividing line of the ath first working surface image and the corresponding adjacent pixel points in the kth second dividing line of the bth second working surface image, that is, the difference in the change direction between the jth first dividing line and the kth second dividing line. The smaller the value of this formula is, the more consistent the change trend between the two dividing lines is, that is, the greater the directional consistency is.

[0097] Finally, the boundary similarity between the first dividing line and the second dividing line is multiplied by the direction consistency between the first dividing line and the second dividing line, so as to obtain the boundary matching degree between the first dividing line and the second dividing line.

[0098] Through this embodiment, the boundary matching degree between the first dividing line and the second dividing line is calculated according to the boundary similarity between the first dividing line and the second dividing line, and the direction consistency between the first dividing line and the second dividing line. In this way, the boundary matching degree between the first dividing line and the second dividing line can be accurately determined by comprehensively considering the consistency of the position and the consistency of the change trend, and the calculation accuracy of the boundary matching degree between the first dividing line and the second dividing line can be improved.

[0099] As an optional embodiment, S301 may specifically include:

[0100] For each edge line in the first working surface image and the second working surface image, the following steps are performed respectively:

[0101] Obtaining the filling performance degree of each adjacent working surface area corresponding to the target edge line, the target edge line being any one of the edge lines;

[0102] The absolute value of the difference between the maximum filling performance degree and the minimum filling performance degree in the filling performance degrees of each adjacent working surface area is multiplied by the maximum filling performance degree to obtain the boundary line probability of the target edge line;

[0103] When the boundary line probability of the target edge line is greater than a preset probability threshold, it is determined that the target edge line belongs to the boundary line.

[0104] In this embodiment, the adjacent working surface area is used to characterize the working surface area that is adjacent to the edge of the target.

[0105] As an example, for each edge line in the first working surface image and the second working surface image, the corresponding boundary line probability is calculated by the following formula 4:

[0106] Formula 4

[0107] In the formula, The boundary line probability used to characterize the jth edge line, It is used to characterize the maximum filling performance degree among the filling performance degrees of each adjacent working surface area corresponding to the jth edge line, It is used to characterize the minimum filling performance degree among the filling performance degrees of adjacent working surface areas corresponding to the j-th edge line.

[0108] in, It is used to characterize the difference between the filling performance levels of adjacent working surface areas corresponding to the jth edge line. When the formula is larger, the contrast is more obvious, and the possibility of the boundary line between the filling body area and other areas is greater, that is, the probability of the boundary line is greater.

[0109] Finally, after obtaining the boundary line probability of each edge line in the first working surface image and the second working surface image, the boundary line probability of each edge line is compared with the preset probability threshold in turn, and each edge line whose boundary line probability is greater than the preset probability threshold is determined as a boundary line.

[0110] Through this embodiment, based on the difference between the filling performance levels of each adjacent working surface area corresponding to the edge line, the probability of the boundary line corresponding to each edge line is calculated, so that the boundary line in each edge line can be accurately screened out. In this way, considering the difference between the filling performance levels of each adjacent working surface area corresponding to the edge line, it is possible to accurately determine whether the edge line belongs to the boundary line, which can improve the accuracy of the boundary line determination.

[0111] As an optional embodiment, S302 may specifically include:

[0112] Obtaining the minimum distance between the first dividing line and the second dividing line;

[0113] The absolute value of the difference between the boundary line probability of the first boundary line and the boundary line probability of the second boundary line is multiplied by the inverse of the minimum distance value to obtain a third calculation result;

[0114] An exponential function operation is performed on the third calculation result to obtain the boundary line similarity between the first boundary line and the second boundary line.

[0115] In this embodiment, the boundary line similarity between the first boundary line and the second boundary line can be specifically determined by the following formula 5:

[0116] Formula 5

[0117] In the formula, Used to characterize the boundary line similarity between the j-th first boundary line of the a-th first working surface image and the k-th second boundary line of the b-th second working surface image. It is used to represent the jth first dividing line of the ath first working surface image, The kth second dividing line used to represent the bth second working surface image, Used to characterize the minimum distance between the j-th first dividing line of the a-th first working surface image and the k-th second dividing line of the b-th second working surface image. The boundary line probability used to characterize the jth first boundary line of the ath first working surface image, It is used to characterize the boundary line probability of the kth second boundary line of the bth second working surface image. exp is used to characterize the exponential function operation with a natural constant as the base.

[0118] Among them, the minimum distance between the jth first dividing line of the ath first working surface image and the kth second dividing line of the bth second working surface image is The smaller it is, the closer the positions of the j-th first dividing line and the k-th second dividing line between two adjacent working surface images are, that is, the greater the boundary similarity between the j-th first dividing line and the k-th second dividing line.

[0119] in, It is used to characterize the difference in boundary line probabilities between the jth first boundary line of the ath first working surface image and the kth second boundary line of the bth second working surface image. The greater the difference in boundary line probabilities, the less similar the jth first boundary line is to the kth second boundary line, that is, the smaller the boundary line similarity between the jth first boundary line and the kth second boundary line is.

[0120] Through this embodiment, the boundary similarity between the first dividing line and the second dividing line is calculated according to the distance between the first dividing line and the second dividing line, and the boundary probability difference between the first dividing line and the second dividing line. In this way, the boundary similarity between the first dividing line and the second dividing line can be accurately calculated by comprehensively considering the distance and the boundary probability difference, thereby improving the calculation accuracy of the boundary similarity.

[0121] As an optional embodiment, Figure 4 As shown, S104 may specifically include the following S401 to S403:

[0122] S401, performing feature matching on the first working surface image and the second working surface image to obtain a plurality of matching point pairs between the first working surface image and the second working surface image and corresponding initial matching degrees;

[0123] S402, determining the reference position consistency of each matching point pair according to the matching degree of each dividing line;

[0124] S403: Based on the consistency of each reference position, correct the initial matching degree of the corresponding matching point pair to obtain the final matching degree of the matching point pair between the first working surface image and the second working surface image.

[0125] In this embodiment, the reference position consistency is used to characterize the similarity between the relative positional relationships between the two points in the matching point pair and the corresponding dividing line.

[0126] As an example, the server uses a feature extraction algorithm (such as SIFT, SURF, ORB, etc.) to extract feature points and their descriptors from the first working surface image and the second working surface image. Among them, feature points are usually prominent points in the image (such as corner points, edge points, etc.), and descriptors describe the image information around these feature points. Use a feature matching algorithm (such as brute force matching, FLANN matching, etc.) to match the feature points in the first working surface image with the feature points in the second working surface image. The purpose of this step is to find similar feature points in the two images, that is, matching point pairs. For each pair of matching points, calculate its initial matching degree. The matching degree is usually determined based on a similarity measure between descriptors (such as Euclidean distance, Hamming distance, etc.). The higher the initial matching degree, the more similar the two feature points are.

[0127] Then, based on the matching degree of each corresponding boundary line in the first working surface image and the second working surface image, the first target boundary line and the second target boundary line with the highest boundary line matching degree are found. According to the relative position relationship between the first matching point of the matching point pair in the first working surface image and the first target boundary line, and the relative position relationship between the second matching point of the matching point pair in the second working surface image and the second target boundary line, the reference position consistency of the matching point pair of the first working surface image and the second working surface image is determined.

[0128] Finally, for all matching point pairs, a weight is set according to the consistency of their corresponding reference positions, and the initial matching degree of the matching point pairs is corrected according to the weight to obtain the final matching degree of the matching point pairs.

[0129] Specifically, the final matching degree of the matching point pair between the first working surface image and the second working surface image can be determined by the following formula 6:

[0130] Formula 6

[0131] In the formula, It is used to characterize the final matching degree of the matching point pair consisting of the cth feature point and the dth feature between the first working surface image and the second working surface image, It is used to characterize the initial matching degree of the matching point pair consisting of the cth feature point and the dth feature between the first working surface image and the second working surface image, Used to characterize the reference position consistency of the matching point pair consisting of the cth feature point and the dth feature between the first working surface image and the second working surface image.

[0132] in, The greater the consistency of the reference positions of the matching point pairs used to characterize the cth feature point and the dth feature between the first working surface image and the second working surface image, the greater the possibility that the cth feature point and the dth feature are the same point, that is, the greater the corresponding final matching degree.

[0133] Through this embodiment, the reference position consistency of the matching point pair is determined according to the matching degree of the dividing line. Then, based on the reference position consistency of the matching point pair, the initial matching degree of the matching point pair is corrected to obtain the final matching degree of the matching point pair. In this way, the matching degree of the matching point pair is corrected based on the similarity of the relative position relationship between the two points in the matching point pair and the corresponding dividing line, so as to improve the accuracy of the final matching degree of the matching point pair.

[0134] As an optional embodiment, the matching point pair includes a first matching point in the first working surface image and a second matching point in the second working surface image;

[0135] S402 may specifically include:

[0136] Based on the matching degree of each dividing line, a first target dividing line corresponding to the maximum dividing line matching degree is selected from each first dividing line, and a second target dividing line corresponding to the maximum dividing line matching degree is selected from each second dividing line;

[0137] Obtaining a first angle between a first connecting line and a first target dividing line, and obtaining a second angle between a second connecting line and a second target dividing line, wherein the first connecting line is a connecting line between a first matching point and a midpoint of the first target dividing line, and the second connecting line is a connecting line between a second matching point and a midpoint of the second target dividing line;

[0138] Based on the first angle, the second angle and the maximum boundary line matching degree, the reference position consistency of the matching point pair is calculated.

[0139] In this embodiment, the first target boundary line is the first boundary line in the first working surface image corresponding to the maximum boundary line matching degree, and the second target boundary line is the second boundary line in the second working surface image corresponding to the maximum boundary line matching degree.

[0140] The first angle is the angle formed by the first connecting line formed by the first matching point in the first working surface image and the midpoint of the first target dividing line and the first target dividing line, and the second angle is the angle formed by the second connecting line formed by the second matching point in the second working surface image and the midpoint of the second target dividing line and the second target dividing line.

[0141] As an example, the server selects the first target boundary line corresponding to the maximum boundary line matching degree for each first boundary line in the first working surface image; and selects the second target boundary line corresponding to the maximum boundary line matching degree for each second boundary line in the second working surface image.

[0142] Then, a first connecting line between the first matching point and the midpoint of the first target dividing line and a second connecting line between the second matching point and the midpoint of the second target dividing line are obtained, and a first angle between the first connecting line and the first target dividing line and a second angle between the second connecting line and the second target dividing line are calculated respectively.

[0143] Finally, based on the first angle, the second angle and the maximum boundary line matching degree, the reference position consistency of the matching point pair is calculated by the following formula 7:

[0144] Formula 7

[0145] In the formula, It is used to characterize the reference position consistency of the matching point pair consisting of the cth feature point and the dth feature between the first working surface image and the second working surface image, It is used to characterize the maximum boundary matching degree between the first working surface image and the second working surface image. It is used to characterize the first angle corresponding to the cth feature point in the first working surface image. It is used to characterize the second angle corresponding to the dth feature point in the second working surface image, and exp is used to characterize the exponential function operation.

[0146] in, It is used to characterize the difference between the first angle corresponding to the cth feature point in the first working surface image and the second angle corresponding to the dth feature point in the second working surface image. The larger the formula is, the more consistent the relative position relationship between the two feature points and the corresponding target dividing line is, and the greater the consistency of the reference position is.

[0147] Through this embodiment, the reference position consistency of the matching point pair is accurately calculated according to the matching degree of each dividing line. In this way, based on the similarity of the relative position relationship between the two points in the matching point pair and the corresponding dividing line, it is helpful to correct the matching degree of the matching point pair, which can improve the accuracy of the final matching degree of the matching point pair.

[0148] As an optional embodiment, S105 may specifically include:

[0149] Based on the final matching degree of each matching point pair, a target matching point pair whose final matching degree is greater than a preset matching degree threshold is screened out;

[0150] Based on the target matching point pair, the first working surface image is aligned with the second working surface image to obtain a target stitching image;

[0151] The depth information of each pixel in the target stitched image is converted into three-dimensional point coordinates to form three-dimensional point cloud data corresponding to the target stitched image;

[0152] Using 3D point cloud data, a digital 3D model of the coal mine filling mining face is constructed.

[0153] In this embodiment, the server selects target matching point pairs whose final matching degrees are greater than a preset matching degree threshold according to the final matching degrees of each matching point pair.

[0154] Then, the transformation matrix (such as affine transformation, perspective transformation, etc.) between the two images is calculated using the target matching point pairs, and then the transformation matrix is ​​applied to align the second working surface image to the first working surface image, and finally the images are stitched together to obtain the target stitched image. Specifically, the image stitching algorithm can be selected from weighted averaging, Laplacian pyramid fusion, etc.

[0155] Then, traverse each pixel in the target stitched image, and calculate the three-dimensional coordinates of the pixel in three-dimensional space based on its depth information and the camera's intrinsic parameters (such as focal length, optical center, etc.) and extrinsic parameters (such as camera position, posture, etc.), and obtain the three-dimensional point cloud data corresponding to the target stitched image.

[0156] Finally, use a 3D point cloud processing library (such as PCL, Open3D, etc.) to preprocess the point cloud data, such as denoising and downsampling; use a meshing algorithm (such as Delaunay triangulation algorithm, Ball-Pivoting algorithm, etc.) to convert the point cloud data into a triangular mesh to obtain a digital 3D model of the coal mine filling mining face; and export the processed 3D model to a common 3D file format (such as OBJ, STL, etc.) for further processing and visualization in 3D modeling software.

[0157] Through this embodiment, based on the final matching degree of the matching point pair, the first working surface image and the second working surface image are aligned to obtain a target spliced ​​image. Finally, according to the target spliced ​​image, a digital three-dimensional model of the coal mine filling mining working surface is constructed. In this way, this embodiment can improve the accuracy of constructing a digital three-dimensional model of the coal mine filling mining working surface by accurately splicing the first working surface image and the second working surface image and then constructing a three-dimensional model.

[0158] It should be clear that the present invention is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present invention is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present invention.

[0159] It should also be noted that the exemplary embodiments mentioned in the present invention describe some methods or systems based on a series of steps or devices. However, the present invention is not limited to the order of the above steps, that is, the steps can be performed in the order mentioned in the embodiments, or in a different order from the embodiments, or several steps can be performed simultaneously.

[0160] The above is only a specific implementation of the present invention. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working process of the system, module and unit described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here. It should be understood that the protection scope of the present invention is not limited to this. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be covered within the protection scope of the present invention.

Claims

1. A method for constructing a digital three-dimensional model of a coal mine filling mining working face, characterized in that: The method comprises: Acquire a first working face image and a second working face image of coal mine backfill mining, wherein the first working face image and the second working face image respectively include a plurality of working face areas and a plurality of edge lines, and the first working face image and the second working face image are images of two adjacent areas in the coal mine backfill mining working face; Determining the filling expression degree of each of the working surface areas in the first working surface image and the second working surface image respectively according to the gray value of each pixel point in the first working surface image and the second working surface image, wherein the filling expression degree is used to characterize the possibility that the working surface area belongs to the filling body area; Determining the degree of boundary matching between a first boundary line of the first working surface image and a second boundary line of the second working surface image according to the filling expression degree of each of the working surface areas, wherein the boundary line is an edge line between the filling body area and other areas among a plurality of edge lines of the working surface image; According to the matching degree of each of the dividing lines, the initial matching degree of the matching point pairs between the first working surface image and the second working surface image is corrected to obtain the final matching degree of the matching point pairs between the first working surface image and the second working surface image; Using the final matching degree of each of the matching point pairs, the first working surface image and the second working surface image are spliced ​​to form a digital three-dimensional model of the coal mine filling mining working surface; Determining the filling expression degree of each working surface area in the first working surface image and the second working surface image respectively according to the gray value of each pixel point in the first working surface image and the second working surface image includes: For each of the working surface areas in the first working surface image and the second working surface image, the following steps are performed respectively: Determine the color contrast of the target working surface area according to the grayscale value of each pixel in the target working surface area, the target working surface area is any one of the working surface areas, and the color contrast is used to characterize the color vividness of the target working surface area; Determining a filling degree of the target working surface area according to a color contrast of the target working surface area and a grayscale value of each pixel point in the target working surface area; Determining the filling degree of the target working surface area according to the color contrast of the target working surface area and the grayscale value of each pixel in the target working surface area includes: Based on the grayscale value of each pixel in the target working surface area, calculating the grayscale value difference between adjacent pixels in the target working surface area; Calculating the average value of each gray value difference to obtain a second calculation result; Multiplying the second calculation result by the color contrast of the target working surface area to obtain the filling performance degree of the target working surface area; Determining the degree of boundary line matching between the first boundary line of the first working surface image and the second boundary line of the second working surface image according to the filling expression degree of each of the working surface areas includes: Based on the filling expression degree of each of the working surface areas, the edge lines in the first working surface image and the second working surface image are screened to obtain a first dividing line in the first working surface image and a second dividing line in the second working surface image; determining a boundary line similarity between the first dividing line and the second dividing line according to a distance between the first dividing line and the second dividing line; Determining the direction consistency of the first dividing line and the second dividing line according to the gradient angle difference between each adjacent pixel point in the first dividing line and the gradient angle difference between each adjacent pixel point in the second dividing line; Based on the boundary similarity and the direction consistency, a boundary matching degree between the first boundary and the second boundary is calculated.

2. The method for constructing a digital three-dimensional model of a coal mine filling mining working face according to claim 1, characterized in that: Determining the color contrast of the target working surface area according to the grayscale value of each pixel in the target working surface area includes: The grayscale value of each pixel in the target working surface area is counted to obtain an average grayscale value, a maximum grayscale value and a minimum grayscale value in the target working surface area; The difference between the average gray value and the minimum gray value is divided by the difference between the maximum gray value and the minimum gray value, and the absolute value is taken to obtain a first calculation result; The first calculation result is multiplied by the average gray value to obtain the color contrast of the target working surface area.

3. The method for constructing a digital three-dimensional model of a coal mine filling mining working face according to claim 1, characterized in that: The step of screening the edge lines in the first working surface image and the second working surface image based on the filling expression degree of each working surface area to obtain a first boundary line in the first working surface image and a second boundary line in the second working surface image includes: For each of the edge lines in the first working surface image and the second working surface image, the following steps are performed respectively: Obtaining the filling performance degree of each adjacent working surface area corresponding to the target edge line, wherein the target edge line is any one of the edge lines; The absolute value of the difference between the maximum filling performance degree and the minimum filling performance degree in the filling performance degrees of each of the adjacent working surface areas is multiplied by the maximum filling performance degree to obtain the boundary line probability of the target edge line; When the boundary line probability of the target edge line is greater than a preset probability threshold, it is determined that the target edge line belongs to the boundary line.

4. The method for constructing a digital three-dimensional model of a coal mine filling mining working face according to claim 3, characterized in that: The determining, according to the distance between the first dividing line and the second dividing line, the dividing line similarity between the first dividing line and the second dividing line comprises: Obtaining a minimum distance between the first dividing line and the second dividing line; The absolute value of the difference between the first dividing line probability and the second dividing line probability is multiplied by the negative number of the minimum distance value to obtain a third calculation result; An exponential function operation is performed on the third calculation result to obtain the boundary line similarity between the first boundary line and the second boundary line.

5. The method for constructing a digital three-dimensional model of a coal mine backfilling working face according to claim 1, characterized in that: The method of correcting the initial matching degree of the matching point pairs between the first working surface image and the second working surface image according to the matching degree of each of the dividing lines to obtain the final matching degree of the matching point pairs between the first working surface image and the second working surface image includes: Performing feature matching on the first working surface image and the second working surface image to obtain a plurality of matching point pairs between the first working surface image and the second working surface image and corresponding initial matching degrees; Determining the consistency of the reference positions of the matching point pairs according to the matching degree of the dividing lines; Based on the consistency of each reference position, the initial matching degree of the corresponding matching point pair is corrected to obtain the final matching degree of the matching point pair between the first working surface image and the second working surface image.

6. The method for constructing a digital three-dimensional model of a coal mine backfilling working face according to claim 5, characterized in that: The matching point pair includes a first matching point in the first working surface image and a second matching point in the second working surface image; Determining the consistency of the reference positions of the matching point pairs according to the matching degree of the dividing lines includes: Based on the matching degrees of each of the dividing lines, a first target dividing line corresponding to the maximum dividing line matching degree is selected from each of the first dividing lines, and a second target dividing line corresponding to the maximum dividing line matching degree is selected from each of the second dividing lines; Obtaining a first angle between a first connecting line and the first target dividing line, and obtaining a second angle between a second connecting line and the second target dividing line, wherein the first connecting line is a connecting line between the first matching point and the midpoint of the first target dividing line, and the second connecting line is a connecting line between the second matching point and the midpoint of the second target dividing line; Based on the first angle, the second angle and the maximum boundary line matching degree, the reference position consistency of the matching point pair is calculated.

7. The method for constructing a digital three-dimensional model of a coal mine backfilling working face according to any one of claims 1 to 6, characterized in that: The method of using the final matching degree of each of the matching point pairs to splice the first working surface image with the second working surface image to form a digital three-dimensional model of the coal mine filling mining working surface includes: Based on the final matching degree of each of the matching point pairs, a target matching point pair whose final matching degree is greater than a preset matching degree threshold is obtained by screening; Based on the target matching point pair, aligning the first working surface image with the second working surface image to obtain a target stitching image; Converting the depth information of each pixel in the target stitched image into three-dimensional point coordinates to form three-dimensional point cloud data corresponding to the target stitched image; The three-dimensional point cloud data is used to construct a digital three-dimensional model of the coal mine filling mining working face.

Images