A dump line generation method and system based on a grid map
By using image processing and edge extraction techniques based on raster maps, continuous and smooth spoil lines are generated, solving the problems of complex generation process and poor robustness in existing technologies, and realizing efficient and automated spoil line generation and path planning.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- 安徽海博智能科技有限责任公司
- Filing Date
- 2026-02-27
- Publication Date
- 2026-06-12
AI Technical Summary
Existing technologies for generating spoil lines in open-pit mine spoil heaps rely on complex topographic surveys and multi-source data fusion, resulting in cumbersome steps, high data update costs, low automation, and poor robustness. They are unable to adapt to the dynamic changes in spoil heaps, affecting operational safety and efficiency.
A grid-based map approach is adopted to repair edge defects through image processing, identify and extract the main area of the dumping line, perform fan-shaped cropping and edge extraction, identify the edge of the retaining wall using convolution operation, and generate a continuous and smooth dumping line by fitting Bézier curves.
It simplifies the process of generating spoil disposal lines, reduces data collection and processing costs, improves operational efficiency and robustness, and allows the generated spoil disposal lines to be dynamically adjusted to adapt to changes in the spoil disposal site, ensuring operational safety and the reliability of path planning.
Smart Images
Figure CN122199736A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of raster map dump line generation technology, and particularly to a method and system for generating dump lines based on raster maps. Background Technology
[0002] In open-pit mine spoil heaps and other operational environments, the efficient and accurate generation of spoil lines is crucial for planning unloading routes, ensuring operational safety, and improving spoil disposal efficiency. Current technologies typically rely on complex topographic surveys, manual design, or multi-source data fusion for spoil line generation, resulting in cumbersome steps, high data update costs, and low automation. Specifically, existing methods often require the collection and processing of large amounts of different types of data (such as elevation point clouds and boundary coordinates), leading to computationally intensive generation processes and difficulty in adapting to dynamic changes in spoil heap terrain as operations progress. Furthermore, when data quality is poor (e.g., missing data, noise, or discontinuities), existing methods exhibit poor robustness, easily leading to interruptions, bends, or insufficient accuracy in the generated spoil lines, thus affecting the reliability of subsequent vehicle path planning and operational safety.
[0003] Therefore, there is an urgent need for a method that can automatically, efficiently, and robustly generate high-precision spoil disposal lines relying solely on a single, readily available data source, in order to simplify the workflow, reduce data acquisition and processing costs, and adapt to the dynamic changes of spoil disposal sites. Summary of the Invention
[0004] The purpose of this invention is to overcome the shortcomings of the existing technology. To achieve the above objective, a method and system for generating dumping lines based on grid maps is adopted to solve the problems mentioned in the background technology.
[0005] A method for generating spoil lines based on raster maps includes the following steps:
[0006] S1. Obtain the raster map data of the spoil heap; S2. Perform image processing on the raster map data to repair edge defects in the map and enhance the continuity of the soil removal line area to obtain a preprocessed raster map. S3. Based on the set initial point pose, identify and extract the main area of the dumping line in the preprocessed grid map; S4. Using the initial point pose as a reference, the main area of the soil discharge line is fan-shaped to obtain the target fan-shaped area; S5. Perform edge extraction processing on the target fan-shaped area to identify the edge of the retaining wall of the spoil heap; S6. Extract and sort the edge point set of the retaining wall of the spoil heap to obtain an ordered edge point sequence; S7. Based on the ordered edge point sequence, perform trajectory fitting to generate the final continuous and smooth soil discharge line.
[0007] As a further aspect of the present invention: In step S2, the image processing includes performing a morphological opening operation on the raster map, specifically as follows: The grid Figure 2 After being valued or converted to grayscale, an erosion operation is first performed to eliminate minor noise and separate weak connections. Then, a dilation operation is performed to restore the shape of the main area and fill the boundary pits. Finally, it is converted back to a raster map to obtain a preprocessed raster map with continuous and smooth edges.
[0008] As a further aspect of the present invention: in step S3, identifying and extracting the main area of the spoil heap specifically includes: Starting from the set point pose (x, y, heading), move along the heading angle with a fixed step size and map to the grid map index; Search for the first grid point whose occupancy value is a preset threshold; If the grid point is found, a breadth-first search algorithm is used to obtain a set of all connected grid points with an occupancy value of the preset threshold, and this set is marked as the main area of the waste disposal line.
[0009] As a further aspect of the present invention: in step S4, the fan-shaped cutting specifically includes: A reference straight line is determined based on the position and pose of the point; Rotate the reference line clockwise and counterclockwise by a preset angle respectively to obtain two boundary lines; Traverse the raster map and retain only the raster points located between the two boundary lines; Adjust the preset angle to determine the cutting boundary that is closest to the actual obstacle, and record the corresponding boundary index point.
[0010] As a further aspect of the present invention: in step S5, the edge extraction processing is implemented using convolution operations, specifically including: Traverse the cropped raster map to obtain the occupancy state array of each raster point and its upper right 2x2 neighborhood; The occupied state array is convolutionally calculated and matched using a predefined set of convolutional kernels; Based on the convolution matching results, grid points that match the edge features of the retaining wall are marked to obtain an edge grid map.
[0011] As a further aspect of the present invention: before step S6, the following is also included: Connectivity analysis is performed on the raster map obtained after edge extraction. The connected region with the largest area is retained as the effective edge of the retaining wall, and noise and discrete edge points are filtered out.
[0012] As a further aspect of the present invention: in step S6, the extraction and sorting of the edge point set specifically includes: Calculate the distance from all edge points to the reference line described in step S4, and find the closest point as the starting point; Based on the position of each point relative to the reference line, it is divided into two sets; The problem of sorting the edge points is transformed into a traveling salesman problem. The points in the two sets are sorted respectively to obtain an ordered sequence that starts from the starting point and visits all edge points in sequence. Based on the boundary index points recorded in step S4, a subset of valid edge points located between two index points is cut out from the ordered sequence.
[0013] As a further aspect of the present invention: in step S7, the trajectory fitting uses a Bezier curve to fit the subset of effective edge points to generate a smooth soil discharge line trajectory.
[0014] As a further aspect of the present invention, the grid map of the spoil heap is also used for planning the unloading path of the spoil heap vehicles.
[0015] The second aspect of the technical solution: An automatic generation system employing a raster map-based spoil line generation method as described in any of the above claims, comprising: The data acquisition module is used to acquire raster map data of the spoil heap; The image preprocessing module is used to perform image processing on the raster map data, repair edge defects, and obtain a preprocessed raster map. The main body recognition module is used to identify and extract the main body area of the dumping line in the preprocessed raster map based on the set initial point pose. The region trimming module is used to perform fan-shaped trimming on the main area of the soil dumping line to obtain the target fan-shaped area; The edge extraction module is used to perform convolution operations on the target fan-shaped region to identify the edge of the retaining wall; The point set processing module is used to extract and sort the edge point sets of the retaining wall of the spoil heap to obtain an ordered sequence of edge points. The trajectory generation module is used to perform curve fitting based on the ordered sequence of edge points to generate the final continuous and smooth soil discharge line.
[0016] Compared with the prior art, the present invention has the following technical advantages: Using the above technical solution, a raster map of the spoil heap is acquired and preprocessed to repair edge defects. Then, based on a preset initial point pose, the main area of the spoil heap line is identified and extracted from the preprocessed map. Next, using this point pose as a reference, the main area is fan-shaped cropped to focus on the target work area. Then, edge extraction techniques such as convolution are used to accurately identify the retaining wall edge from the cropped area. The extracted discrete edge points are then sorted, filtered, and optimized to form an ordered point sequence. Finally, a continuous and smooth spoil heap line trajectory is generated by curve fitting of this point sequence. This method significantly simplifies the process of generating dump lines, reduces reliance on complex data collection and manual design, thereby greatly improving operational efficiency and reducing costs. Furthermore, the method is insensitive to the quality of the input raster map, exhibiting good robustness and generating continuous and reliable dump lines even when the map has local defects or noise. In addition, the dump lines generated by this method can be dynamically adjusted as the raster map is updated, ensuring the real-time performance and adaptability of dump operations. Moreover, the same map can be directly used for vehicle routing planning, achieving "one map, multiple uses." Attached Figure Description
[0017] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings: Figure 1 This is a schematic diagram illustrating the steps of the method for generating a waste disposal line according to an embodiment of this application; Figure 2 This is a flowchart illustrating the method for generating a waste disposal line according to an embodiment of this application. Figure 3 This is a schematic diagram of the initial point pose of an embodiment disclosed in this application; Figure 4 This is a schematic diagram of the final generated waste disposal line trajectory according to an embodiment disclosed in this application. Detailed Implementation
[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0019] Please refer to Figure 1 and Figure 2 In this embodiment of the invention, a method for generating dump lines based on a raster map includes the following steps: S1. Obtain the raster map data of the spoil heap; S2. Perform image processing on the raster map data to repair edge defects in the map and enhance the continuity of the soil removal line area to obtain a preprocessed raster map. In step S2, the image processing includes performing morphological opening operations on the raster map, specifically as follows: The grid Figure 2 After being valued or converted to grayscale, an erosion operation is first performed to eliminate minor noise and separate weak connections. Then, a dilation operation is performed to restore the shape of the main area and fill the boundary pits. Finally, it is converted back to a raster map to obtain a preprocessed raster map with continuous and smooth edges.
[0020] In the specific implementation steps, the original raster map of the spoil heap is first preprocessed to address issues such as local missing parts, depressions, or discontinuities at its edges. Specifically, the opening operation method from morphological image processing is used: the raster map is first converted to a grayscale image, then erosion and dilation operations are performed sequentially, and finally converted back to a raster map. This processing effectively fills small holes and smooths uneven boundaries, resulting in a continuous, complete raster map with smooth edges, laying the foundation for accurate extraction of the main area of the spoil heap line.
[0021] S3. Based on the set initial point pose, identify and extract the main area of the dumping line in the preprocessed grid map; In step S3, identifying and extracting the main area of the waste dumping line specifically includes: Starting from the set point pose (x, y, heading), move along the heading angle with a fixed step size and map to the grid map index; Search for the first grid point whose occupancy value is a preset threshold; If the grid point is found, a breadth-first search algorithm is used to obtain a set of all connected grid points with an occupancy value of the preset threshold, and this set is marked as the main area of the waste disposal line.
[0022] In the specific implementation steps, step S3 (disposal line entity identification) aims to accurately extract continuous areas belonging to the disposal line from the preprocessed raster map. Specifically, it includes: like Figure 3 As shown in the figure, the diagram illustrates the initial point pose. The points and arrows in the diagram represent the preset initial point poses (including coordinate positions and heading directions) in the spoil heap grid map, which are used to guide the identification and fan-shaped clipping of the main spoil heap area.
[0023] Based on the preset initial point pose (including position coordinates x, y and heading angle), a progressive search is performed along the heading direction with a fixed step size until a grid point with a preset obstacle marker (e.g., 100) is first retrieved in the grid map. If the search exceeds the map boundary, the identification is deemed to have failed. If the grid point is successfully located, it is used as a seed, and a breadth-first search algorithm is used to obtain all grid points with the same marker connected to it. These grid points are marked as the main area of the spoil line, and the remaining grid points are set to non-main markers (e.g., 0), thus forming a continuous and complete spoil line area.
[0024] S4. Using the initial point pose as a reference, the main area of the soil discharge line is fan-shaped to obtain the target fan-shaped area; In step S4, the fan-shaped cutting specifically includes: A reference straight line is determined based on the position and pose of the point; Rotate the reference line clockwise and counterclockwise by a preset angle respectively to obtain two boundary lines; Traverse the raster map and retain only the raster points located between the two boundary lines; Adjust the preset angle to determine the cutting boundary that is closest to the actual obstacle, and record the corresponding boundary index point.
[0025] In the specific implementation steps, step S4 (dumping grid trimming) is used to constrain the spatial range of the main area of the identified dumping line in order to extract the effective dumping line segment related to the current operation direction. The specific implementation method is as follows: Based on the point pose set in S2, a reference straight line is determined; this straight line is rotated clockwise and counterclockwise by a certain initial angle (e.g., 89°) to obtain two boundary straight lines, which together form a fan-shaped area; the grid map is traversed, and only the grid points located within this fan-shaped area are retained, while the rest are set to invalid values. Further, by dynamically adjusting the rotation angle of the above two boundary straight lines, the fan-shaped range that best fits the actual obstacle distribution is found, and the key grid indices (first_index and second_index) corresponding to the two boundaries of this range are recorded for subsequent edge extraction and trajectory generation.
[0026] S5. Perform edge extraction processing on the target fan-shaped area to identify the edge of the retaining wall of the spoil heap; In step S5, the edge extraction process is implemented using convolution operations, specifically including: Traverse the cropped raster map to obtain the occupancy state array of each raster point and its upper right 2x2 neighborhood; The occupied state array is convolutionally calculated and matched using a predefined set of convolutional kernels; Based on the convolution matching results, grid points that match the edge features of the retaining wall are marked to obtain an edge grid map.
[0027] In the specific implementation steps, step S5 (extraction of spoil line edge points) is used to extract a valid and ordered set of spoil line contour points from the edge raster map, which specifically includes the following processing: First, to remove noise and discrete edges, a connected component analysis is performed on the raster map after edge extraction: all occupied rasters are traversed, depth-first search is used to identify all interconnected occupied regions, the area of each connected region is calculated, and the region with the largest area is selected; then a new raster map is generated, retaining only the occupied state of the largest connected region, thus obtaining the denoised single-connected spoil heap edge raster map.
[0028] In the specific implementation steps, grid edge extraction is performed; To extract the edges of the retaining wall, a convolution operation needs to be performed on the raster map. The overall process of the convolution operation is as follows: 1) Traverse the cropped raster map in S3 and obtain the occupancy state array of the upper right 2x2 field of the current raster element; 2) Use a predefined kernel group (Core) to perform convolution operations on the surrounding values of the current grid; 3) Update the corresponding grid values in the map based on the convolution results, thereby realizing the edge extraction of the retaining wall.
[0029] The specific computational details of the convolution operation are as follows: traverse the grid map and calculate the array representing the occupied state for each grid point. The subscripts 0-3 represent the four grid cells above, to the upper right, itself, and to the right of the current grid cell, respectively. The calculation method is as follows:
[0030] To achieve edge detection, the convolutional kernel group used contains 9 convolutional kernels, denoted as the convolutional kernel group. Let the k-th convolutional kernel be... The input array is Then the matching score of the k-th convolutional kernel for:
[0031] in, It's the Kronecker Delta function, which then finds the value relative to the current raster array A. For a fully matched convolutional kernel, first calculate the maximum score `maxValue` and its index `maxIndex`, respectively:
[0032]
[0033] If the result is found This indicates that the convolution kernel perfectly matches the current grid and belongs to the edge of the retaining wall. Therefore, the data of the four points in the upper right corner of the grid are all updated to 100. By traversing the entire grid map, the retaining wall boundary grid map is obtained.
[0034] S6. Extract and sort the edge point set of the retaining wall of the spoil heap to obtain an ordered edge point sequence; In step S6, the extraction and sorting of the edge point set specifically includes: Perform connectivity analysis on the grid obtained by edge extraction in step S5, filter out noise and discrete points, and retain the connected region with the largest area as the effective edge of the retaining wall. Calculate the distance from all edge points to the reference line described in step S4, and find the closest point as the starting point; Based on the position of each point relative to the reference line, it is divided into two sets; The problem of sorting the edge points is transformed into a traveling salesman problem. The points in the two sets are sorted respectively to obtain an ordered sequence that starts from the starting point and visits all edge points in sequence. Based on the boundary index points recorded in step S4, a subset of valid edge points located between two index points is cut out from the ordered sequence.
[0035] In the specific implementation steps, based on the above technical solution, the further processing of the edge raster map and the generation of ordered point sets are as follows: First, to remove redundant edges and noise interference, a connected component analysis needs to be performed on the raster map obtained after edge extraction. A depth-first search is used to traverse all occupied rasters, identifying all interconnected obstacle rasters, and calculating and comparing the areas of each connected component. The connected component with the largest area is retained, generating a new edge raster map containing only the occupied state of that region, thus achieving edge filtering and purification.
[0036] Building upon this foundation, a geometric sorting mechanism is introduced to construct an ordered sequence of edge points suitable for subsequent trajectory fitting. Based on the reference line defined by the point vectors determined in S2, the vertical distance from each occupied point in the new edge raster map to this line is calculated, and a sorting starting point is determined according to the principle of closest proximity. Simultaneously, based on the orientation of each point relative to the reference line, it is divided into two subsets (e.g., Side A and Side B). By modeling the connection order of each set of points as a traveling salesman problem and solving it, an ordered sequence of edge points along the runoff line is finally obtained, providing structured input data for subsequent trajectory fitting.
[0037] S7. Based on the ordered edge point sequence, perform trajectory fitting to generate the final continuous and smooth soil discharge line.
[0038] In step S7, the trajectory fitting uses a Bézier curve to fit the subset of effective edge points to generate a smooth soil discharge line trajectory.
[0039] like Figure 4 As shown in the figure, this is a schematic diagram of the final generated spoil heap trajectory. The curve in the figure represents the continuous and smooth spoil heap trajectory generated after identifying, sorting, and fitting the edge points using the method of this invention. In the specific steps, based on the key boundary index determined by S3, effective segments are extracted from the ordered edge point sequence and the trajectory is smoothed. Specifically, this includes: converting the index points first_index and second_index into actual coordinate points left_point and right_point in the raster map; then, finding the points in the ordered point set origiPoints that have the closest Euclidean distance to left_point and right_point respectively, obtaining the indices left_index and right_index, and extracting a subset of points between these two points to accurately extract the effective edge of the retaining wall near the vehicle operation side. Finally, a Bézier curve is used to fit this subset of points to generate a continuous and smooth waste disposal line trajectory, serving as the direct basis for path planning.
[0040] In this embodiment, the grid map of the spoil heap is also used for planning the unloading path of the spoil heap vehicles.
[0041] The second aspect of the technical solution: An automatic generation system employing a raster map-based spoil line generation method as described in any of the above claims, comprising: The data acquisition module is used to acquire raster map data of the spoil heap; The image preprocessing module is used to perform image processing on the raster map data, repair edge defects, and obtain a preprocessed raster map. The main body recognition module is used to identify and extract the main body area of the dumping line in the preprocessed raster map based on the set initial point pose. The region trimming module is used to perform fan-shaped trimming on the main area of the soil dumping line to obtain the target fan-shaped area; The edge extraction module is used to perform convolution operations on the target fan-shaped region to identify the edge of the retaining wall; The point set processing module is used to extract and sort the edge point sets of the retaining wall of the spoil heap to obtain an ordered sequence of edge points. The trajectory generation module is used to perform curve fitting based on the ordered sequence of edge points to generate the final continuous and smooth soil discharge line.
[0042] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention. The scope of the invention is defined by the appended claims and their equivalents, all of which should be included within the scope of protection of the invention.
Claims
1. A method for generating spoil lines based on raster maps, characterized in that, Includes the following steps: S1. Obtain the raster map data of the spoil heap; S2. Perform image processing on the raster map data to repair edge defects in the map and enhance the continuity of the soil removal line area to obtain a preprocessed raster map. S3. Based on the set initial point pose, identify and extract the main area of the dumping line in the preprocessed grid map; S4. Using the initial point pose as a reference, the main area of the soil discharge line is fan-shaped to obtain the target fan-shaped area; S5. Perform edge extraction processing on the target fan-shaped area to identify the edge of the retaining wall of the spoil heap; S6. Extract and sort the edge point set of the retaining wall of the spoil heap to obtain an ordered edge point sequence; S7. Based on the ordered edge point sequence, perform trajectory fitting to generate the final continuous and smooth soil discharge line.
2. The method for generating spoil lines based on a raster map according to claim 1, characterized in that, In step S2, the image processing includes performing morphological opening operations on the raster map, specifically as follows: After binarizing or converting the raster map to a grayscale image, an erosion operation is first performed to eliminate minor noise and separate weak connections. Then, a dilation operation is performed to restore the shape of the main area and fill the boundary pits. Finally, the image is converted back to a raster map to obtain a preprocessed raster map with continuous and smooth edges.
3. The method for generating spoil lines based on a raster map according to claim 1, characterized in that, In step S3, identifying and extracting the main area of the spoil heap specifically includes: Starting from the set point pose (x, y, heading), move along the heading angle with a fixed step size and map to the grid map index; Search for the first grid point whose occupancy value is a preset threshold; If the grid point is found, a breadth-first search algorithm is used to obtain a set of all connected grid points with an occupancy value of the preset threshold, and this set is marked as the main area of the waste disposal line.
4. The method for generating spoil lines based on a raster map according to claim 1, characterized in that, In step S4, the fan-shaped cutting specifically includes: A reference straight line is determined based on the position and pose of the point; Rotate the reference line clockwise and counterclockwise by a preset angle respectively to obtain two boundary lines; Traverse the raster map and retain only the raster points located between the two boundary lines; Adjust the preset angle to determine the cutting boundary that is closest to the actual obstacle, and record the corresponding boundary index point.
5. The method for generating spoil lines based on a raster map according to claim 1, characterized in that, In step S5, the edge extraction process is implemented using convolution operations, specifically including: Traverse the cropped raster map to obtain the occupancy state array of each raster point and its upper right 2x2 neighborhood; The occupied state array is convolutionally calculated and matched using a predefined set of convolutional kernels; Based on the convolution matching results, grid points that match the edge features of the retaining wall are marked to obtain an edge grid map.
6. The method for generating dump lines based on a raster map according to claim 5, characterized in that, The steps preceding step S6 also include: Connectivity analysis is performed on the raster map obtained after edge extraction. The connected region with the largest area is retained as the effective edge of the retaining wall, and noise and discrete edge points are filtered out.
7. The method for generating spoil lines based on a raster map according to claim 1, characterized in that, In step S6, the extraction and sorting of the edge point set specifically includes: Calculate the distance from all edge points to the reference line described in step S4, and find the closest point as the starting point; Based on the position of each point relative to the reference line, it is divided into two sets; The problem of sorting the edge points is transformed into a traveling salesman problem. The points in the two sets are sorted respectively to obtain an ordered sequence that starts from the starting point and visits all edge points in sequence. Based on the boundary index points recorded in step S4, a subset of valid edge points located between two index points is cut out from the ordered sequence.
8. The method for generating spoil lines based on a raster map according to claim 7, characterized in that, In step S7, the trajectory fitting uses a Bézier curve to fit the subset of effective edge points to generate a smooth soil discharge line trajectory.
9. The method for generating spoil lines based on a raster map according to claim 1, characterized in that, The grid map of the spoil heap is also used for planning the unloading routes of spoil heap vehicles.
10. An automatic generation system employing the method for generating spoil lines based on a raster map as described in any one of claims 1 to 9, characterized in that, include: The data acquisition module is used to acquire raster map data of the spoil heap; The image preprocessing module is used to perform image processing on the raster map data, repair edge defects, and obtain a preprocessed raster map. The main body recognition module is used to identify and extract the main body area of the dumping line in the preprocessed raster map based on the set initial point pose. The region trimming module is used to perform fan-shaped trimming on the main area of the soil dumping line to obtain the target fan-shaped area; The edge extraction module is used to perform convolution operations on the target fan-shaped region to identify the edge of the retaining wall; The point set processing module is used to extract and sort the edge point sets of the retaining wall of the spoil heap to obtain an ordered sequence of edge points. The trajectory generation module is used to perform curve fitting based on the ordered sequence of edge points to generate the final continuous and smooth soil discharge line.