Ramp detection method, processor and ramp detection apparatus for a mechanical device

Abstract

Embodiments of the present application provide a slope detection method for a mechanical device, a processor, a machine readable storage medium, a slope detection device and a mechanical device, the slope detection method comprising: obtaining first three-dimensional point cloud data of a working environment of the mechanical device; constructing a point cloud digital elevation model according to the first three-dimensional point cloud data, the point cloud digital elevation model comprising a plurality of first grids and an elevation value corresponding to each first grid; determining an elevation distribution feature according to the first three-dimensional point cloud data; determining a target grid according to the elevation distribution feature and the elevation value, the target grid being a point cloud grid associated with a slope; and determining a detection result of the slope according to point clouds in the point cloud grid associated with the slope, the slope detection method being capable of quickly and accurately identifying a slope in the working environment of the mechanical device, and the method being simple and easy to implement.

Description

Technical Field

[0001] This invention relates to the field of mechanical equipment technology, and more specifically to a slope detection method, processor, machine-readable storage medium, slope detection device, and mechanical equipment for use in mechanical equipment. Background Technology

[0002] Unmanned driving and intelligent operation of machinery (such as excavators) are current industry technology hotspots. For machinery operating scenarios, environmental perception capabilities (such as slope perception) are required to detect and digitally analyze various targets (such as obstacles, slopes, and ditches) during the machinery's movement and operation. Currently, non-real-time environmental perception of machinery is mainly achieved through GPS-based 3D guidance systems. Real-time environmental perception is primarily achieved through sensors such as LiDAR and depth cameras. When using depth cameras to perceive the slope environment of machinery, it is necessary to first acquire ground frame point clouds, then mesh and perform mean filtering on the point clouds, and finally segment the slope area using gradient thresholding. However, this method is mainly for uneven ground, and single thresholding cannot accurately identify slope point clouds. When using LiDAR to perceive the slope environment of machinery, slope detection is achieved through planar and linear fitting, which is also for uneven ground and slopes, making accurate slope extraction difficult. Summary of the Invention

[0003] The purpose of this invention is to provide a slope detection method, processor, machine-readable storage medium, slope detection device, and mechanical equipment for mechanical equipment. This slope detection method, processor, machine-readable storage medium, slope detection device, and mechanical equipment can quickly and accurately identify slopes in the working environment of mechanical equipment, and the method is simple and easy to implement.

[0004] To achieve the above objectives, a first aspect of the present invention provides a slope detection method for mechanical equipment, comprising:

[0005] Acquire first-dimensional point cloud data of the operating environment of the mechanical equipment;

[0006] A point cloud digital elevation model is constructed based on the first three-dimensional point cloud data. The point cloud digital elevation model includes multiple first grids and elevation values ​​corresponding to each first grid.

[0007] Determine the elevation distribution characteristics based on the first three-dimensional point cloud data;

[0008] The target grid is determined based on the elevation distribution characteristics and elevation values. The target grid is a point cloud grid associated with the slope.

[0009] The detection results of the slope are determined based on the point cloud in the point cloud grid associated with the slope.

[0010] In an embodiment of the present invention, the first three-dimensional point cloud data is the original three-dimensional point cloud data of the working environment of the mechanical equipment, or the three-dimensional point cloud data after voxel meshing processing of the original three-dimensional point cloud data.

[0011] In an embodiment of the present invention, constructing a point cloud digital elevation model based on first three-dimensional point cloud data includes:

[0012] Construct a first grid based on the first three-dimensional point cloud data;

[0013] The first three-dimensional point cloud data is projected onto the first grid based on the first and second coordinate values ​​of the first three-dimensional point cloud data to construct a point cloud digital elevation model, wherein the first and second coordinate values ​​are coordinate values ​​on a plane perpendicular to the elevation direction.

[0014] In an embodiment of the present invention, constructing a first mesh based on first three-dimensional point cloud data includes:

[0015] Determine the maximum and minimum first coordinate values ​​of the first 3D point cloud data;

[0016] The number of columns in the first grid is determined based on the maximum and minimum first coordinate values;

[0017] Determine the maximum and minimum second coordinate values ​​of the first 3D point cloud data;

[0018] The number of rows in the first grid is determined based on the maximum and minimum second coordinate values;

[0019] Construct the first grid based on the number of columns and rows.

[0020] In an embodiment of the present invention, the elevation distribution feature includes at least one of ground relief and elevation dispersion, wherein ground relief is the range of elevation values ​​of the point cloud in the first grid; and elevation dispersion is the average deviation of elevation values ​​of the point cloud in the first grid.

[0021] In embodiments of the present invention, elevation distribution characteristics include ground relief and elevation dispersion, and determining the target grid based on elevation distribution characteristics and elevation values ​​includes:

[0022] A weighted summation operation is performed on the ground relief and elevation dispersion to determine the characteristic value of the elevation distribution.

[0023] The first grid in the point cloud digital elevation model is divided according to the elevation distribution characteristic values ​​to determine the second grid;

[0024] The target grid is determined based on the elevation value and the second grid.

[0025] In an embodiment of the present invention, the first grid in the point cloud digital elevation model is divided according to the elevation distribution characteristic value to determine the second grid, which includes:

[0026] Obtain the preset elevation distribution segmentation threshold;

[0027] The first grid is divided into a ground point cloud grid and a non-ground point cloud grid based on the elevation distribution characteristic value and the preset elevation distribution segmentation threshold.

[0028] The non-ground point cloud grid was identified as the second grid.

[0029] In an embodiment of the present invention, determining the target grid based on the elevation value and the second grid includes:

[0030] The slope value is determined based on the elevation values ​​of adjacent grids in the second grid.

[0031] Obtain the preset slope segmentation threshold;

[0032] The second grid is divided according to the slope value and the preset slope segmentation threshold to determine the target grid.

[0033] In embodiments of the present invention, determining the detection result of the slope based on the point cloud in the point cloud mesh associated with the slope includes:

[0034] Clustering algorithms are used to cluster the point clouds in the target grid to determine the set of point clouds after clustering.

[0035] The slope's outline and / or orientation are determined based on the clustered point cloud set.

[0036] A second aspect of the present invention provides a processor configured to perform the above-described method for mechanical devices.

[0037] A third aspect of the present invention provides a machine-readable storage medium storing instructions that cause a machine to perform the above-described slope detection method for mechanical equipment.

[0038] A fourth aspect of the present invention provides a slope detection device for mechanical equipment, comprising:

[0039] The data acquisition module is used to acquire the first three-dimensional point cloud data of the operating environment of the mechanical equipment;

[0040] The model building module is used to build a point cloud digital elevation model based on the first three-dimensional point cloud data. The point cloud digital elevation model includes multiple first grids and elevation values ​​corresponding to each first grid.

[0041] The feature determination module is used to determine the elevation distribution features based on the first three-dimensional point cloud data.

[0042] The target grid determination module is used to determine the target grid based on the elevation distribution characteristics and elevation values. The target grid is a point cloud grid associated with the slope.

[0043] The slope detection module is used to determine the slope detection result based on the point cloud in the point cloud mesh associated with the slope.

[0044] The fifth aspect of the present invention provides a mechanical device, the mechanical device comprising:

[0045] A scanning device for scanning the operating environment of mechanical equipment to generate first three-dimensional point cloud data; and

[0046] The processor mentioned above.

[0047] The above technical solution acquires first-dimensional point cloud data of the operating environment of mechanical equipment, constructs a point cloud digital elevation model based on the first-dimensional point cloud data and determines the elevation distribution characteristics, determines the point cloud grid associated with the slope based on the elevation distribution characteristics and the elevation values ​​in the point cloud digital elevation model, and finally determines the slope detection result based on the point cloud grid. This slope detection method for mechanical equipment can quickly and accurately identify slopes in the operating environment of mechanical equipment, providing a source of perception information for the automatic driving and operation of mechanical equipment.

[0048] Other features and advantages of the embodiments of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0049] The accompanying drawings are provided to further illustrate embodiments of the present invention and form part of the specification. They are used together with the following detailed description to explain the embodiments of the present invention, but do not constitute a limitation thereof. In the drawings:

[0050] Figure 1 This is a schematic diagram of the first process of the slope detection method for mechanical equipment in an embodiment of the present invention;

[0051] Figure 2 This is a schematic diagram of the second process of the slope detection method for mechanical equipment in an embodiment of the present invention;

[0052] Figure 3 This is a schematic diagram of the composition of the slope detection device in an embodiment of the present invention.

[0053] Explanation of reference numerals in the attached figures

[0054] 1. Data Acquisition Module 2. Model Building Module

[0055] 3 Feature Determination Module 4 Target Mesh Determination Module

[0056] 5. Slope Detection Module Detailed Implementation

[0057] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the scope of the present invention.

[0058] Embodiments of the present invention provide a slope detection method for mechanical equipment, such as... Figure 1 As shown, the slope detection method includes the following steps:

[0059] Step S101: Obtain the first three-dimensional point cloud data of the working environment of the mechanical equipment.

[0060] Specifically, the mechanical equipment in this embodiment can be an excavator, which includes a scanning device and a processor that is communicatively connected to the scanning device. Further, the scanning device can be a lidar sensor installed on the excavator body, which can scan the excavator's working environment in real time. The processor includes a data acquisition module 1. After the scanning device completes the scanning, it sends the generated scanning data to the processor so that the data acquisition module 1 in the processor can acquire the first three-dimensional point cloud data.

[0061] Step S102: Construct a point cloud digital elevation model based on the first three-dimensional point cloud data. The point cloud digital elevation model includes multiple first grids and elevation values ​​corresponding to each first grid.

[0062] In one embodiment of the present invention, step S102: constructing a point cloud digital elevation model based on the first three-dimensional point cloud data includes the following steps:

[0063] Construct a first grid based on the first three-dimensional point cloud data;

[0064] The first three-dimensional point cloud data is projected onto the first grid based on the first and second coordinate values ​​of the first three-dimensional point cloud data to construct a point cloud digital elevation model, wherein the first and second coordinate values ​​are coordinate values ​​on a plane perpendicular to the elevation direction.

[0065] In one embodiment of the present invention, the first three-dimensional point cloud data is the original three-dimensional point cloud data of the working environment of the mechanical equipment, or the three-dimensional point cloud data after voxel meshing processing of the original three-dimensional point cloud data.

[0066] The data directly obtained by the scanning device after scanning is the original three-dimensional point cloud data of the working environment of the mechanical equipment. If the original three-dimensional point cloud data is transmitted to the processor without processing and directly sent to the data acquisition module 1, then the first three-dimensional point cloud data obtained by the data acquisition module 1 is the original three-dimensional point cloud data of the working environment of the mechanical equipment. In another embodiment of the present invention, the processor also includes a voxel meshing processing module, which can perform voxel meshing processing on the original three-dimensional point cloud. In this embodiment, the scanning device first transmits the original three-dimensional point cloud data to the voxel meshing processing module, which performs voxel meshing processing on the original three-dimensional point cloud data. The data obtained after voxel meshing processing is the first three-dimensional point cloud data. The voxel meshing processing module then sends the first three-dimensional point cloud data to the data acquisition module 1. This method does not lose the point cloud contour features of the working environment of the mechanical equipment, and can reduce the amount of point cloud data processing and improve the efficiency of data processing.

[0067] Specifically, the voxel mesh processing module, based on the concept of voxel meshes, performs voxel meshing on the original 3D point cloud data. The voxel meshed original 3D point cloud data is transformed into a collection of multiple tiny 3D cubes (i.e., a voxel mesh set). Further, voxel meshing of the original 3D point cloud data requires first setting a fixed voxel mesh side length l. o (in this embodiment l) o (Pre-stored in the processor and retrieved when needed), then the maximum and minimum values ​​of the above data in the x, y, and z directions are obtained based on the original 3D point cloud data, and l is obtained. o After determining the maximum and minimum values ​​of the above data in the x, y, and z directions, the overall resolution a*b*c of the voxel mesh set is determined according to the following formula:

[0068]

[0069]

[0070]

[0071] Where, x max The maximum value of the original 3D point cloud data in the x-direction, x min The minimum value of the original 3D point cloud data in the x-direction, y max The value of the original 3D point cloud data in the y-direction is y. min Let z be the minimum value of the original 3D point cloud data in the y-direction. max The maximum value of the original 3D point cloud data in the z-direction is z. min This represents the minimum value of the original 3D point cloud data in the z-direction.

[0072] After determining the overall resolution a*b*c of the voxel mesh set, the voxel mesh set can be obtained. The original 3D point cloud data is then projected onto the voxel mesh set, and the number of points n in each voxel mesh in the voxel mesh set can be obtained. p The coordinates of the point cloud in the voxel mesh are (x... i ,yi,z i ).

[0073] The 3D point cloud data after voxel meshing is determined using the following formula:

[0074]

[0075]

[0076]

[0077] Among them, C x C represents the x-coordinate of the point cloud in the voxel mesh. y C represents the y-coordinate value of the point cloud in the voxel mesh. z Let n be the z-coordinate of the point cloud in the voxel mesh. p ∑x represents the number of point clouds in the voxel mesh. i x is the x-axis of the point cloud in the voxel mesh. i The sum of values, ∑y i For the point cloud in the voxel mesh, y i The sum of values, ∑z i z-axis of point cloud in voxel mesh i The sum of the values. The point cloud data obtained after voxel meshing according to the above formula is the first three-dimensional point cloud data.

[0078] In one embodiment of the present invention, constructing a first mesh based on the first three-dimensional point cloud data further includes the following steps:

[0079] Determine the maximum and minimum first coordinate values ​​of the first 3D point cloud data;

[0080] The number of columns in the first grid is determined based on the maximum and minimum first coordinate values;

[0081] Determine the maximum and minimum second coordinate values ​​of the first 3D point cloud data;

[0082] The number of rows in the first grid is determined based on the maximum and minimum second coordinate values;

[0083] Construct the first grid based on the number of columns and rows.

[0084] Specifically, the processor also includes a model building module 2. After obtaining the first three-dimensional point cloud data, the data acquisition module 1 sends it to the model building module 2 so that the model building module can construct a point cloud digital elevation model based on the first three-dimensional point cloud data. In this embodiment, the first three-dimensional point cloud data is the three-dimensional point cloud data obtained after voxel meshing of the original three-dimensional point cloud data. According to formulas (4)-(6), the model building module 2 calculates the point cloud (C) of each voxel mesh in the voxel mesh set. x C y C z In this embodiment, the first grid is a two-dimensional regular grid located on the XOY coordinate plane. When constructing the first grid, the side length `grid_size` must first be set (in this embodiment, `grid_size` is pre-stored in the processor and retrieved when needed). Based on the point cloud (C...) of each voxel grid... x C y C z Determine the maximum value C of the point cloud in the x-direction within the voxel mesh set. xmax Minimum value C xmin The maximum value C in the y direction ymax Minimum value C ymin Then, calculate the number of rows and columns of the first grid using the following formula:

[0085]

[0086]

[0087] Where row is the number of rows in the first grid, col is the number of columns in the first grid, and C xmax C represents the maximum value of the point cloud in the x-direction within the voxel mesh set. xmin Let C be the minimum value of the point cloud in the x-direction within the voxel mesh set. ymax C represents the maximum value of the point cloud in the voxel mesh set in the y-direction. ymin It represents the minimum value of the point cloud in the y-direction within the voxel mesh set.

[0088] The point cloud of each voxel mesh (C x C y C z The first coordinate value C x Second coordinate value C y All points are projected into the first grid that has been constructed, thus completing the construction of the point cloud digital elevation model.

[0089] Step S103: Determine the elevation distribution characteristics based on the first three-dimensional point cloud data.

[0090] In one embodiment of the present invention, the elevation distribution feature includes at least one of ground relief and elevation dispersion, wherein ground relief is the range of height values ​​of the point cloud in the first grid; elevation dispersion is the average deviation of height values ​​of the point cloud in the first grid; furthermore, in this embodiment, the processor also includes a feature determination module 3, and the determination of the elevation distribution feature is performed in the feature determination module 3.

[0091] Step S104: Determine the target grid based on the elevation distribution characteristics and elevation values. The target grid is a point cloud grid associated with the slope.

[0092] The processor in this embodiment also includes a target mesh determination module 4, in which the target mesh is determined.

[0093] In one embodiment of the present invention, step S104: the elevation distribution characteristics include ground relief and elevation dispersion, and determining the target grid based on the elevation distribution characteristics and elevation values ​​further includes the following steps:

[0094] A weighted summation operation is performed on the ground relief and elevation dispersion to determine the characteristic value of the elevation distribution.

[0095] The first grid in the point cloud digital elevation model is divided according to the elevation distribution characteristic values ​​to determine the second grid;

[0096] The target grid is determined based on the elevation value and the second grid.

[0097] Specifically, in the case of uneven ground and slopes in excavator operation scenarios, existing technologies extract slopes by segmenting them using a single slope threshold when sensing or detecting slopes. This easily leads to the identification of some uneven ground as slopes, resulting in low slope detection accuracy. However, this embodiment is based on the characteristics of relatively uniform distribution and low dispersion of ground point clouds. It performs statistical analysis on the height values ​​of the first three-dimensional point cloud data and uses multiple statistical features to divide the first grid in the point cloud digital elevation model so as to subsequently determine the ground point cloud grid and the non-ground point cloud grid. More preferably, the multiple statistical feature values ​​in this embodiment include ground undulation and elevation dispersion. Ground undulation is represented by the range of height values ​​of the point cloud in the first grid, and elevation dispersion is represented by the average deviation of height values ​​of the point cloud in the first grid. Ground undulation can be calculated by formula (9), and elevation dispersion can be calculated by formula (10).

[0098] range = C zmax -C zmin (9)

[0099] Where range represents the ground relief, C zmax C represents the maximum height value of the point cloud in the first grid. zminThis represents the minimum height value of the point cloud in the first grid. The ground relief corresponding to each grid in the first grid can be calculated using the formula above.

[0100]

[0101] Where MD (mean deviation) is the elevation dispersion, and C zk C represents the height value of the k-th point cloud in each grid of the first grid (i.e., the value of the point cloud in the z-direction). zmean The value is the average height of all point clouds in each grid of the first grid, and point_size is the total number of point clouds in each grid of the first grid.

[0102] After calculating the elevation dispersion of each grid in the first grid according to the above formula, the ground relief weight and elevation dispersion weight can be calculated based on the following formula:

[0103]

[0104] Among them, w r This represents the weighting of ground undulation.

[0105]

[0106] Among them, w md This represents the elevation dispersion weight.

[0107] The aforementioned weights for ground undulation and elevation dispersion are both adaptive weights. That is, neither of these two weight values ​​is a fixed value, but rather they can adaptively change according to the different working environments of the excavator and the different initial 3D point cloud data generated by the scanning equipment. This is beneficial to further improve the accuracy of ground point cloud and non-ground point cloud segmentation in subsequent steps, thereby improving the accuracy of slope detection results.

[0108] After calculating the ground relief, elevation dispersion, ground relief weight, and elevation dispersion weight, the elevation distribution characteristic value can be calculated using the following formula:

[0109] DS=w r *range+w md *MD (13)

[0110] Among them, DS (distribution statistics) represents the elevation distribution characteristic value.

[0111] In one embodiment of the present invention, dividing the first grid in the point cloud digital elevation model according to the elevation distribution characteristic value to determine the second grid includes the following steps:

[0112] Obtain the preset elevation distribution segmentation threshold;

[0113] The first grid is divided into a ground point cloud grid and a non-ground point cloud grid based on the elevation distribution characteristic value and the preset elevation distribution segmentation threshold.

[0114] The non-ground point cloud grid was identified as the second grid.

[0115] The preset elevation distribution segmentation threshold T in this embodiment DS It is pre-stored in the processor and can be retrieved when needed.

[0116] Feature determination module 3 calculates the elevation distribution feature value DS of each grid in the first grid of the point cloud digital elevation model, and then sends the elevation distribution feature value DS to the target grid determination module 4. The target grid determination module 4 compares the elevation distribution feature value DS of each grid with the preset elevation distribution segmentation threshold T. DS Compare, if DS <T DS If DS > T, then the first grid being compared is determined to be the ground point cloud grid; DS If the first grid being compared is determined to be a non-ground point cloud grid, then the first grid is determined to be a non-ground point cloud grid. In this embodiment, the first grid is divided into ground point cloud grids and non-ground point cloud grids based on the elevation distribution feature value and the preset elevation distribution segmentation threshold. If there is uneven ground in the excavator's working scene, this division method can filter out the interference of ground point cloud in the ground point cloud grid on subsequent slope detection, thereby improving the overall accuracy of slope detection.

[0117] In one embodiment of the present invention, determining the target grid based on the elevation value and the second grid further includes the following steps:

[0118] The slope value is determined based on the elevation values ​​of adjacent grids in the second grid.

[0119] Obtain the preset slope segmentation threshold;

[0120] The second grid is divided according to the slope value and the preset slope segmentation threshold to determine the target grid.

[0121] Specifically, the target mesh determination module 4 first extracts the 3D point cloud data corresponding to the second mesh, and then determines the total number of point clouds in each mesh of the second mesh and the sum of the third coordinate values ​​of the point clouds therein (i.e., the C coordinates of the point clouds in each mesh of the second mesh). z The sum of all points in the second grid is used to calculate the average height of all point clouds in each grid cell. This average height is the elevation value corresponding to each grid cell in the second grid.

[0122]

[0123] Where height is the elevation value of a certain grid in the second grid; ∑C z The sum of the third coordinates of the point cloud of a certain grid in the second grid; point_size is the total number of point clouds of a certain grid in the second grid.

[0124] After calculating the elevation value of a certain grid in the second grid, the target grid determination module 4 can further calculate the slope value based on the elevation value. Specifically, based on the direction pointed to by the positive coordinate axis X of the lidar sensor, the difference between the elevation value of a certain grid (such as the m-th grid of the second grid) in the point cloud digital elevation model and the elevation value of the previous grid (such as the (m-1)-th grid of the second grid) is calculated according to formula (16), and this difference is set as the slope value of the m-th grid in the second grid:

[0125] slope = height m -height m-1 (15)

[0126] Where slope is the slope value of the m-th grid in the second grid, and height... m The elevation value of the m-th grid in the second grid is height. m-1 Let be the elevation value of the (m-1)th grid in the second grid. The slope value of each grid in the second grid can be calculated using the formula above.

[0127] The preset slope segmentation threshold T in this embodiment slope It is pre-stored in the processor and can be retrieved when needed.

[0128] The target grid determination module 4 calculates the slope value of each grid in the second grid, and then sequentially compares the slope value of each grid in the second grid with the preset slope segmentation threshold T. slope Compare, if slope > T slope If the slope is not found, then the compared grid in the second grid is determined to belong to the target grid; if the slope is not found, then the comparison grid in the second grid is determined to belong to the target grid. <T slope If the comparison is not complete, the grid being compared in the second grid is determined to be not the target grid. After all grids in the second grid have been compared one by one, the target grid can be determined. The target grid is the point cloud grid associated with the slope.

[0129] Step S105: Determine the detection result of the slope based on the point cloud in the point cloud grid associated with the slope.

[0130] In this embodiment, the processor also includes a slope detection module 5, and the determination of the slope detection result is performed in the slope detection module 5.

[0131] In one embodiment of the present invention, step S105: determining the detection result of the slope based on the point cloud in the point cloud mesh associated with the slope further includes the following steps:

[0132] Clustering algorithms are used to cluster the point clouds in the target grid to determine the set of point clouds after clustering.

[0133] The slope's outline and / or orientation are determined based on the clustered point cloud set.

[0134] Specifically, after acquiring the target grid, clustering algorithms (such as K-means clustering, Mean-Shift clustering, DBSCAN clustering, etc.) are used to cluster the point clouds in the target grid. After clustering, multiple point cloud clusters C are obtained, each containing a different number of point clouds. After clustering, the slope detection module 5 compares the number of point clouds in each point cloud cluster C with a preset minimum clustering point cloud number threshold T. cluster_number (In this embodiment, the preset minimum clustering point cloud number threshold T) cluster_number (Pre-stored in the processor, it can be retrieved when needed for comparison.) If the number of points in the compared point cloud cluster C exceeds the preset minimum clustering point cloud number threshold T, the comparison is performed. cluster_number If the current point cloud cluster C is not more than the preset minimum clustering point cloud number threshold T, then the next point cloud cluster C will be processed. cluster_number If the current point cloud cluster C being compared is not output, the next point cloud cluster C can be processed directly. In this embodiment, clustering is performed on each point cloud in the target grid, which can filter out discrete and few interfering point clouds in the initial slope point cloud, thus further improving the accuracy of slope detection.

[0135] After obtaining the clustered point cloud set, the slope detection module 5 can draw the outline of the slope based on the three-dimensional coordinate values ​​of each point cloud in the point cloud set. In addition, since each point cloud in the point cloud set has directional information, the slope detection module 5 can determine the location of the slope based on the directional information.

[0136] In another embodiment of the present invention, the slope detection module 5 uses a planar rectangular bounding box to represent the contour of the slope, and the orientation of the planar rectangular bounding box represents the orientation of the slope. Specifically, the slope detection module 5 first obtains the three-dimensional coordinate values ​​of the top left and bottom right point clouds in the slope point cloud set. The three-dimensional coordinate value of the top left point cloud is (x... top_left y top_left , z top_left The three-dimensional coordinates of the bottom right point cloud are (x...).bottom_right y bottom_right , z bottom_right Obtain the midpoint of the two point clouds and use the midpoint relative to the location of the excavator as the orientation of the slope; project the top left and bottom right point clouds onto the XOY plane and draw a planar rectangular bounding box based on their coordinates on the XOY plane, and use the drawn planar rectangular bounding box as the outline of the slope.

[0137] In another embodiment of the invention, the processor performs as follows: Figure 2 The slope detection method shown in the figure specifically involves first acquiring first three-dimensional point cloud data, wherein the first three-dimensional point cloud data is the original three-dimensional point cloud data of the working environment of the mechanical equipment; then performing voxel meshing on the first three-dimensional point cloud data to reduce the amount of point cloud data processing; after voxel meshing, constructing a two-dimensional regular mesh based on the point cloud data obtained after voxel meshing, and further constructing a point cloud digital elevation model based on the constructed two-dimensional regular mesh, which can be used to calculate elevation and slope values, and the point cloud digital elevation model includes multiple first meshes, where the first mesh is a two-dimensional regular mesh; after the point cloud digital elevation model is constructed, multi-scale slope point cloud extraction should be performed, which includes: (1) extracting non-ground point cloud meshes according to multi-scale adaptive weights, wherein the multi-scale adaptive weights include ground relief weights and elevation dispersion weights, and non-ground point cloud meshes are extracted according to multi-scale adaptive weights. The ground point cloud grid is extracted by segmenting a two-dimensional regular grid according to multi-scale adaptive weights; (2) the target grid is extracted according to the preset slope segmentation threshold, wherein the preset slope segmentation threshold is stored in the processor in advance and can be retrieved when used; the processor can obtain the target grid by segmenting and extracting the non-ground point cloud grid according to the preset slope segmentation threshold, and the target grid is the point cloud grid associated with the slope; (3) the interference point cloud is filtered out according to the preset minimum cluster point cloud quantity threshold, wherein the preset minimum cluster point cloud quantity threshold is stored in the processor in advance and can be retrieved when used; the processor filters out the initial slope point cloud in the target grid according to the preset minimum cluster point cloud quantity threshold to remove the discrete and fewer interference point clouds in the initial slope point cloud; finally, the processor determines the slope detection result and outputs it according to the slope point cloud obtained after filtering out the interference point cloud, wherein the slope detection result includes at least the slope location and the slope outline.

[0138] Another embodiment of the present invention provides a processor configured to perform the above-described slope detection method for mechanical equipment.

[0139] Another embodiment of the present invention provides a machine-readable storage medium storing instructions for causing a machine to perform the above-described slope detection method for mechanical equipment.

[0140] like Figure 3 As shown, another embodiment of the present invention provides a slope detection device for mechanical equipment, comprising:

[0141] Data acquisition module 1 is used to acquire the first three-dimensional point cloud data of the working environment of the mechanical equipment;

[0142] Model building module 2 is used to build a point cloud digital elevation model based on the first three-dimensional point cloud data. The point cloud digital elevation model includes multiple first grids and elevation values ​​corresponding to each first grid.

[0143] Feature determination module 3 is used to determine elevation distribution features based on the first three-dimensional point cloud data;

[0144] Target mesh determination module 4 is used to determine the target mesh based on elevation distribution characteristics and elevation values. The target mesh is a point cloud mesh associated with the slope.

[0145] The slope detection module 5 is used to determine the slope detection result based on the point cloud in the point cloud grid associated with the slope.

[0146] In another embodiment of the present invention, the model building module 2 is further configured to build a first grid based on the first three-dimensional point cloud data; and to project the first three-dimensional point cloud data into the first grid based on the first coordinate value and the second coordinate value of the first three-dimensional point cloud data to build a point cloud digital elevation model, wherein the first coordinate value and the second coordinate value are coordinate values ​​on a plane perpendicular to the elevation direction.

[0147] In another embodiment of the present invention, the model building module 2 is further configured to determine the maximum first coordinate value and the minimum first coordinate value of the first three-dimensional point cloud data; determine the number of columns of the first grid based on the maximum first coordinate value and the minimum first coordinate value; determine the maximum second coordinate value and the minimum second coordinate value of the first three-dimensional point cloud data; determine the number of rows of the first grid based on the maximum second coordinate value and the minimum second coordinate value; and construct the first grid based on the number of columns and the number of rows.

[0148] In another embodiment of the present invention, the target grid determination module 4 is further configured to perform a weighted summation operation on the ground relief and elevation dispersion to determine the elevation distribution characteristic value; divide the first grid in the point cloud digital elevation model according to the elevation distribution characteristic value to determine the second grid; and determine the target grid according to the elevation value and the second grid.

[0149] In another embodiment of the present invention, the target grid determination module 4 is further configured to obtain a preset elevation distribution segmentation threshold; divide the first grid into a ground point cloud grid and a non-ground point cloud grid according to the elevation distribution feature value and the preset elevation distribution segmentation threshold; and determine the non-ground point cloud grid as the second grid.

[0150] In another embodiment of the present invention, the target grid determination module 4 is further configured to determine the slope value based on the elevation values ​​of adjacent grids in the second grid; obtain a preset slope segmentation threshold; and divide the second grid according to the slope value and the preset slope segmentation threshold to determine the target grid.

[0151] In another embodiment of the present invention, the slope detection module 5 is further configured to perform clustering processing on each point cloud in the target grid based on a clustering algorithm to determine the clustered point cloud set; and to determine the contour and / or orientation of the slope based on the clustered point cloud set.

[0152] Another embodiment of the present invention provides a mechanical device comprising:

[0153] A scanning device for scanning the operating environment of mechanical equipment to generate first three-dimensional point cloud data; and the processor in the above embodiments.

[0154] This invention provides a slope detection method, processor, machine-readable storage medium, slope detection device, and mechanical equipment for use in machinery. The method acquires first three-dimensional point cloud data of the operating environment of the mechanical equipment, constructs a point cloud digital elevation model based on the first three-dimensional point cloud data, determines the elevation distribution characteristics, determines the point cloud grid associated with the slope based on the elevation distribution characteristics and the elevation values ​​in the point cloud digital elevation model, and finally determines the slope detection result based on the point cloud grid. This slope detection method for mechanical equipment can quickly and accurately identify slopes in the operating environment of mechanical equipment, providing a source of perception information for the automatic driving and operation of mechanical equipment.

[0155] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0156] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0157] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0158] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0159] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0160] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, like read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0161] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0162] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0163] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for slope detection in mechanical equipment, characterized in that, include: Acquire first-dimensional point cloud data of the operating environment of the mechanical equipment; A point cloud digital elevation model is constructed based on the first three-dimensional point cloud data. The point cloud digital elevation model includes multiple first grids and elevation values ​​corresponding to each first grid. The elevation distribution characteristics are determined based on the first three-dimensional point cloud data, wherein the elevation distribution characteristics include ground relief and elevation dispersion, wherein the ground relief is the range of height values ​​of the point cloud in the first grid; and the elevation dispersion is the average deviation of height values ​​of the point cloud in the first grid. A weighted summation operation is performed on the ground relief and the elevation dispersion to determine the elevation distribution characteristic value; The first grid in the point cloud digital elevation model is divided according to the elevation distribution characteristic value to determine the second grid; The target grid is determined based on the elevation value and the second grid, wherein the target grid is a point cloud grid associated with the slope; The detection result of the slope is determined based on the point cloud in the point cloud grid associated with the slope.

2. The slope detection method for mechanical equipment according to claim 1, characterized in that, The first three-dimensional point cloud data is the original three-dimensional point cloud data of the working environment of the mechanical equipment, or the three-dimensional point cloud data after voxel meshing processing of the original three-dimensional point cloud data.

3. The slope detection method for mechanical equipment according to claim 1, characterized in that, The step of constructing a point cloud digital elevation model based on the first three-dimensional point cloud data includes: The first mesh is constructed based on the first three-dimensional point cloud data; The first three-dimensional point cloud data is projected onto the first grid based on the first coordinate value and the second coordinate value of the first three-dimensional point cloud data to construct the point cloud digital elevation model, wherein the first coordinate value and the second coordinate value are coordinate values ​​on a plane perpendicular to the elevation direction.

4. The slope detection method for mechanical equipment according to claim 3, characterized in that, The step of constructing the first mesh based on the first three-dimensional point cloud data includes: Determine the maximum and minimum first coordinate values ​​of the first three-dimensional point cloud data; The number of columns in the first grid is determined based on the maximum first coordinate value and the minimum first coordinate value; Determine the maximum and minimum second coordinate values ​​of the first three-dimensional point cloud data; The number of rows in the first grid is determined based on the maximum second coordinate value and the minimum second coordinate value; The first grid is constructed based on the number of columns and the number of rows.

5. The slope detection method for mechanical equipment according to claim 1, characterized in that, The step of dividing the first grid in the point cloud digital elevation model according to the elevation distribution characteristic value to determine the second grid includes: Obtain the preset elevation distribution segmentation threshold; The first grid is divided into a ground point cloud grid and a non-ground point cloud grid based on the elevation distribution characteristic value and the preset elevation distribution segmentation threshold. The non-ground point cloud grid is identified as the second grid.

6. The slope detection method for mechanical equipment according to claim 5, characterized in that, The step of determining the target grid based on the elevation value and the second grid includes: The slope value is determined based on the elevation values ​​of adjacent grids in the second grid. Obtain the preset slope segmentation threshold; The second grid is divided according to the slope value and the preset slope segmentation threshold to determine the target grid.

7. The slope detection method for mechanical equipment according to claim 1, characterized in that, The step of determining the slope detection result based on the point cloud in the point cloud mesh associated with the slope includes: Clustering algorithms are used to cluster the point clouds in the target grid to determine the clustered point cloud set. The contour and / or orientation of the slope are determined based on the point cloud set after clustering.

8. A processor, characterized in that, It is configured to perform the slope detection method for mechanical equipment according to any one of claims 1 to 7.

9. A machine-readable storage medium storing instructions thereon, characterized in that, The instructions are used to cause the machine to perform the slope detection method for mechanical equipment according to any one of claims 1 to 7.

10. A slope detection device for mechanical equipment, characterized in that, include: The data acquisition module (1) is used to acquire the first three-dimensional point cloud data of the working environment of the mechanical equipment; Model building module (2) is used to build a point cloud digital elevation model based on the first three-dimensional point cloud data. The point cloud digital elevation model includes multiple first grids and elevation values ​​corresponding to each first grid. The feature determination module (3) is used to determine the elevation distribution features based on the first three-dimensional point cloud data, wherein the elevation distribution features include ground relief and elevation dispersion, wherein the ground relief is the range of height values ​​of the point cloud in the first grid; and the elevation dispersion is the average deviation of height values ​​of the point cloud in the first grid. The target grid determination module (4) is used to perform a weighted summation operation on the ground relief and the elevation dispersion to determine the elevation distribution characteristic value; to divide the first grid in the point cloud digital elevation model according to the elevation distribution characteristic value to determine the second grid; and to determine the target grid according to the elevation value and the second grid, wherein the target grid is a point cloud grid associated with the slope. The slope detection module (5) is used to determine the slope detection result based on the point cloud in the point cloud grid associated with the slope.

11. A mechanical device, characterized in that, The mechanical equipment includes: A scanning device for scanning the operating environment of mechanical equipment to generate first three-dimensional point cloud data; and The processor according to claim 8.

Images