Mining area road surface accumulated water detection method, device and equipment and storage medium

Abstract

The embodiment of the invention provides a mining area road surface accumulated water detection method. The method comprises the steps of obtaining point cloud data located in a first coordinate system; mapping the point cloud data to a second coordinate system, and constructing a feature image of the point cloud data according to the coordinate information of each three-dimensional point; through a preset segmentation model, obtaining a road boundary of the strip mine area road from the feature image; according to the coordinate information and the reflection intensity of each three-dimensional point, the ponding area is determined from the first closed area defined by the road boundary, and the problems that in the prior art, the accuracy of pavement ponding detection of the strip mine area is not high, and the automation degree is low are solved.

Description

Technical Field

[0001] The present application relates to the field of autonomous driving in mining areas, and in particular to a method, device, equipment and storage medium for detecting water accumulation on roads in mining areas. Background Art

[0002] In the field of autonomous driving, comprehensive, accurate and real-time perception of obstacles such as accumulated water is a key technology, which is crucial to decision-making, control modules and safety in autonomous driving. Relying on a single sensor for perception is a big challenge. LiDAR can model the surrounding environment in 3D point cloud. The amount of radar beam determines the hardware cost and the sparseness of point cloud. Its measurement range can reach 200m. As the distance increases, the point cloud becomes more sparse. Visual cameras can collect information with rich colors and textures. The perception accuracy of obstacles is related to the focal length and the density of optical imaging pixels. It cannot directly perceive the distance depth information. Solving the depth information with the help of assumptions such as distance calculation often has the problem of large calculation amount and low confidence. At the same time, it is easily affected by the environment. For example, rainy and foggy weather and light incident will cause the image quality to decline, which in turn causes the perception confidence to decrease. Multi-sensor fusion improves the stability and robustness of the algorithm. It has good detection accuracy under complex environmental conditions such as rain and fog, strong light, weak light, and wind and sand. Multi-sensor and multi-feature fusion better guarantees the stability of perception redundancy and accuracy.

[0003] In open-pit mines, the vehicle speed is relatively low, the traffic volume is small, and the transportation tasks and routes are usually highly repetitive, which is very suitable for autonomous driving applications. However, unlike structured roads, open-pit mines are typical unstructured roads with complex terrain, harsh environment, and irregular road boundaries. When vehicles drive in flooded areas with a small friction coefficient, their tires will slip, posing a great safety hazard. Therefore, it is crucial to solve the problems of low accuracy and low degree of automation in the existing technology of detecting flooded roads in open-pit mines. Summary of the invention

[0004] The embodiments of the present application provide a method for detecting waterlogging on roads in mining areas, an electronic device and a storage medium, so as to solve the problems of low accuracy and low degree of automation in detecting waterlogging on roads in mining areas in the prior art.

[0005] The embodiments of the present application provide a method, device, equipment and storage medium for detecting water accumulation on mining roads, which are used to solve the problem that the color contrast of water accumulation on the road is not obvious due to poor lighting, weather and other environments, resulting in a decrease in image quality, thereby affecting the accuracy of detection.

[0006] In a first aspect, an embodiment of the present application provides a method for detecting waterlogging on a mining road, which is applied to an intelligent driving vehicle. The method includes:

[0007] Acquire point cloud data in a first coordinate system; wherein the point cloud data includes: coordinate information and reflection intensity of three-dimensional points of the open-pit mine road; the first coordinate system refers to a coordinate system with the laser radar as the coordinate origin;

[0008] Mapping the point cloud data to a second coordinate system, and constructing a feature image of the point cloud data based on the coordinate information of each three-dimensional point; wherein the second coordinate system refers to a coordinate system with the center of the rear axle of the vehicle body as the coordinate origin; and the feature image is used to indicate the geometric features of the point cloud data in the second coordinate system;

[0009] The road boundary of the open-pit mine road is obtained from the feature image through a preset segmentation model; wherein the segmentation model is obtained through training of historical feature images;

[0010] According to the coordinate information and reflection intensity of each three-dimensional point, the water accumulation area is determined from the first closed area surrounded by the road boundary.

[0011] Optionally, in the method shown above, a grid matrix is ​​preset in the second coordinate system, and the grid matrix includes a plurality of first grids;

[0012] Map the point cloud data to the second coordinate system, and construct a feature image of the point cloud data based on the coordinate information of each 3D point, including:

[0013] Map point cloud data to a raster matrix;

[0014] Divide the plurality of three-dimensional points based on the plurality of first grids to obtain a point cloud cluster of each first grid;

[0015] According to the point cloud cluster of each first grid and the coordinate information of each three-dimensional point, a grid feature matrix of the grid matrix is ​​obtained;

[0016] Convert the raster feature matrix into a feature image of point cloud data.

[0017] Optionally, the method shown above obtains a grid feature matrix of the grid matrix according to the point cloud cluster of each first grid and the coordinate information of each three-dimensional point, including:

[0018] According to the point cloud cluster of each first grid and the coordinate information of each three-dimensional point, the geometric features of each first grid are obtained; wherein the geometric features include: maximum height, average height and height difference;

[0019] The geometric features of each first grid are fused to obtain a grid feature matrix of the grid matrix.

[0020] Optionally, the method shown above obtains the road boundary of the open-pit mine road from the feature image through a preset segmentation model, including:

[0021] The feature image is input into the segmentation model to obtain the initial boundary of the open-pit mine road output by the segmentation model;

[0022] The second closed area enclosed by the initial boundary of the road is segmented cyclically, and the road boundary is obtained according to the segmentation result.

[0023] Optionally, as shown in the above method, the second closed area enclosed by the initial boundary of the road is segmented cyclically, and the road boundary is obtained according to the segmentation result, including:

[0024] The second closed area is segmented cyclically by an inscribed circle algorithm until the diameter of the largest inscribed circle segmented is smaller than a preset diameter;

[0025] The road boundary is obtained according to the points of at least one largest inscribed circle.

[0026] Optionally, the method shown above, according to the coordinate information and reflection intensity of each three-dimensional point, determines the waterlogged area from the first closed area surrounded by the road boundary, including:

[0027] According to the reflection intensity of each three-dimensional point, a plurality of target points are screened out from the plurality of three-dimensional points; wherein each target point is a three-dimensional point located in the first closed area, and the reflection intensity of each target point is less than a preset intensity;

[0028] Determine an area to be detected from the first closed area according to the multiple target points;

[0029] Rasterizing the area to be detected, dividing multiple target points based on multiple second grids of the area to be detected, and obtaining a point cloud cluster of each second grid;

[0030] According to the point cloud cluster of each second grid, as well as the coordinate information and reflection intensity of each target point, a water accumulation area is determined from the first closed area.

[0031] Optionally, as in the method shown above, before mapping the point cloud data to the second coordinate system and constructing a feature image of the point cloud data according to the coordinate information of each three-dimensional point, the method further includes:

[0032] Preprocessing the coordinate information of each three-dimensional point; wherein the preprocessing includes denoising, correcting geometric errors and normalization;

[0033] If the closed area is a two-dimensional feature image, the waterlogged area is determined from the first closed area surrounded by the road boundary according to the coordinate information and reflection intensity of each three-dimensional point. The method further includes:

[0034] The two-dimensional feature image is upgraded to obtain the water accumulation area detection result in the three-dimensional scene.

[0035] In a second aspect, an embodiment of the present application provides a device for detecting water on a road in a mining area, which is applied to an intelligent driving vehicle, and the device includes:

[0036] An acquisition module is used to acquire point cloud data located in a first coordinate system; wherein the point cloud data includes: coordinate information and reflection intensity of three-dimensional points of the open-pit mine road; the first coordinate system refers to a coordinate system with the laser radar as the coordinate origin;

[0037] A mapping module, used to map the point cloud data to a second coordinate system, and construct a feature image of the point cloud data according to the coordinate information of each three-dimensional point; wherein the second coordinate system refers to a coordinate system with the center of the rear axle of the vehicle body as the coordinate origin; and the feature image is used to indicate the geometric features of the point cloud data in the second coordinate system;

[0038] A segmentation module is used to obtain the road boundary of the open-pit mine road from the feature image through a preset segmentation model; wherein the segmentation model is obtained by training the historical feature image;

[0039] The determination module is used to determine the water accumulation area from the first closed area surrounded by the road boundary according to the coordinate information and reflection intensity of each three-dimensional point.

[0040] Optionally, the mapping module of the device as shown above includes:

[0041] A first mapping module, used for mapping point cloud data to a grid matrix;

[0042] A division module, used for dividing a plurality of three-dimensional points based on a plurality of first grids to obtain a point cloud cluster of each first grid;

[0043] A first calculation module, used for obtaining a grid feature matrix of the grid matrix according to the point cloud cluster of each first grid and the coordinate information of each three-dimensional point;

[0044] The conversion module is used to convert the raster feature matrix into a feature image of point cloud data.

[0045] Optionally, in the device shown above, the first computing module includes:

[0046] A second calculation module is used to obtain the geometric features of each first grid according to the point cloud cluster of each first grid and the coordinate information of each three-dimensional point; wherein the geometric features include: maximum height, average height and height difference;

[0047] The fusion module is used to fuse the geometric features of each first grid to obtain a grid feature matrix of the grid matrix.

[0048] Optionally, the device as shown above, the segmentation module, includes:

[0049] The first segmentation module is used to input the feature image into the segmentation model to obtain the initial boundary of the open-pit mine road output by the segmentation model;

[0050] The first cyclic segmentation module is used to perform cyclic segmentation on the second closed area enclosed by the initial boundary of the road, and obtain the road boundary according to the segmentation result.

[0051] Optionally, in the device shown above, the first cycle segmentation module includes:

[0052] A second cyclic segmentation module is used to perform cyclic segmentation on the second closed area by using an inscribed circle algorithm until the diameter of the largest inscribed circle obtained by segmentation is smaller than a preset diameter;

[0053] The collection module is used to obtain the road boundary according to the points of at least one maximum inscribed circle.

[0054] Optionally, the device as shown above, the determination module includes:

[0055] A screening module, used to screen out a plurality of target points from a plurality of three-dimensional points according to the reflection intensity of each three-dimensional point; wherein each target point is a three-dimensional point located in the first closed area, and the reflection intensity of each target point is less than a preset intensity;

[0056] A first determination module, used to determine the area to be detected from the first closed area according to the multiple target points;

[0057] A second division module is used to rasterize the area to be detected, divide multiple target points based on multiple second grids of the area to be detected, and obtain a point cloud cluster of each second grid;

[0058] The second determination module is used to determine the water accumulation area from the first closed area according to the point cloud cluster of each second grid and the coordinate information and reflection intensity of each target point.

[0059] Optionally, the mapping module of the device shown above further includes:

[0060] A preprocessing module, used for preprocessing the coordinate information of each three-dimensional point; wherein the preprocessing includes denoising, correcting geometric errors and normalization;

[0061] The closed area is a two-dimensional feature image, and the determination module also includes:

[0062] The image processing module is used to upgrade the two-dimensional feature image to obtain the water accumulation area detection result in the three-dimensional scene.

[0063] A third aspect of an embodiment of the present invention provides an electronic device, including: a memory and a processor;

[0064] Memory stores computer-executable instructions;

[0065] The processor executes computer-executable instructions stored in the memory to implement the method for detecting waterlogging on mining roads of any one of the first aspects.

[0066] A fourth aspect of an embodiment of the present invention provides a computer-readable storage medium, in which computer execution instructions are stored. When the computer execution instructions are executed by a processor, they are used to implement the method for detecting waterlogging on mining roads of any one of the first aspects.

[0067] A fifth aspect of an embodiment of the present invention provides a computer program product, including a computer program. When the computer program is executed by a processor, it is used to implement the method for detecting waterlogging on mining roads of any one of the first aspects.

[0068] The mining road waterlogging detection method, electronic device and storage medium provided in the embodiments of the present application obtain point cloud data located in a first coordinate system; wherein the point cloud data includes: coordinate information and reflection intensity of three-dimensional points of the open-pit mining road; the first coordinate system refers to a coordinate system with a laser radar as the coordinate origin; the point cloud data is mapped to a second coordinate system, and a feature image of the point cloud data is constructed based on the coordinate information of each three-dimensional point; wherein the second coordinate system refers to a coordinate system with the center of the rear axle of the vehicle body as the coordinate origin; the feature image is used to indicate the geometric features of the point cloud data in the second coordinate system; through a preset segmentation model, the road boundary of the open-pit mining road is obtained from the feature image; wherein the segmentation model is obtained by training with historical feature images; according to the coordinate information and reflection intensity of each three-dimensional point, the waterlogged area is determined from the first closed area surrounded by the road boundary. The following technical effects are achieved: raster conversion is performed by acquiring multi-frame laser point cloud information; each grid unit is processed according to the vehicle positioning information, and coordinate conversion is performed after processing; when the first water surface area is not greater than the preset first threshold, the corresponding contour is filtered out; according to the remaining contour, the position of the road surface water in the first coordinate system is determined; segmentation is performed according to the semantic segmentation annotation of the water accumulation to obtain the segmentation result, which is fused as the water accumulation recognition result, and the point cloud convex hull contour is recognized to obtain the water accumulation recognition result; the road boundary features of the three-dimensional point cloud data are extracted by a deep learning method, and the road boundary features are fitted using a road model-based method to obtain the road boundary; the reflectivity information of the point cloud is used to distinguish between the water accumulation area and the non-water accumulation area, so as to realize the detection of road surface water accumulation. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0070] Figure 1 Schematic diagram of the process of detecting waterlogging on mining roads provided by the embodiment of the present invention Figure 1 ;

[0071] Figure 2 Schematic diagram of the process of detecting waterlogging on mining roads provided by the embodiment of the present invention Figure 2 ;

[0072] Figure 3 Schematic diagram of the process of detecting waterlogging on mining roads provided by the embodiment of the present invention Figure 3 ;

[0073] Figure 4 Schematic diagram of the process of detecting waterlogging on mining roads provided by the embodiment of the present invention Figure 4 ;

[0074] Figure 5 Schematic diagram of the process of detecting waterlogging on mining roads provided by the embodiment of the present invention Figure 5 ;

[0075] Figure 6 A schematic diagram of the structure of a device for detecting waterlogging on mining roads provided by an embodiment of the present invention;

[0076] Figure 7 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention.

[0077] The above drawings have shown clear embodiments of the present application, which will be described in more detail later. These drawings and text descriptions are not intended to limit the scope of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments.

[0078] Reference numerals:

[0079] 610 - acquisition module; 620 - mapping module; 630 - segmentation module; 640 - determination module; 710 - processor; 720 - memory; 730 - communication component; 740 - bus. DETAILED DESCRIPTION

[0080] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0081] The technical solution of the present invention is described in detail with specific embodiments below. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present invention will be described below in conjunction with the accompanying drawings.

[0082] In the embodiments of the present invention, words such as "first" and "second" are used to distinguish between identical or similar items with substantially the same functions and effects. Those skilled in the art will understand that words such as "first" and "second" do not limit the quantity and execution order, and words such as "first" and "second" do not necessarily limit the difference. It should be noted that in the embodiments of the present invention, words such as "exemplary" or "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "exemplary" or "for example" in this application should not be interpreted as being more preferred or more advantageous than other embodiments or design. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way. In the embodiments of the present invention, "at least one" refers to one or more, and "more" refers to two or more.

[0083] It should be noted that the "at..." in the embodiment of the present invention can be the instant when a certain situation occurs, or it can be a period of time after the situation occurs, and the embodiment of the present invention does not specifically limit this. In addition, the method for detecting waterlogging on mining roads provided in the embodiment of the present invention is only used as an example, and the method for detecting waterlogging on mining roads can also include more or less content.

[0084] In order to clearly describe the technical solutions of the embodiments of the present invention, some terms and technologies involved in the embodiments of the present invention are briefly introduced below:

[0085] Point cloud data: is a digital representation of an object or environment in three-dimensional space, consisting of a large number of points, each of which contains coordinates (X, Y, Z) and possible other attributes such as color, intensity, classification, etc.

[0086] Raster matrix: A method of representing spatial data by a matrix of pixels (or pixels) organized in rows and columns (or grids). Each pixel contains a value that can represent information such as image reflectivity, temperature, etc.

[0087] Point cloud cluster: In three-dimensional space, through cluster analysis of point cloud data, a collection of similar points in space is formed. Point cloud data is a data set composed of many points in three-dimensional space. These points are usually represented by three-dimensional coordinates and sometimes contain attributes such as color and intensity.

[0088] In order to clearly understand the technical solution of the present application, the solution of the prior art is first introduced in detail.

[0089] Road waterlogging detection based on machine vision: Use a camera installed on a vehicle to obtain image data of roads in the mining area, and detect road waterlogging through image processing algorithms. This method is low-cost, but has certain requirements for environmental conditions such as lighting and weather. At the same time, the color contrast of road waterlogging is not obvious, which leads to a decrease in image quality, thereby affecting the accuracy of detection.

[0090] LiDAR-based road waterlogging detection: LiDAR technology is used to obtain ground point cloud data. LiDAR is not affected by factors such as light and weather, and can provide stable position and shape information under various conditions. LiDAR-based road waterlogging detection can be divided into traditional point cloud processing methods and deep learning-based methods. The mining area has complex scenes and poor road flatness, which is not suitable for traditional point cloud processing methods. The deep learning-based method requires strong data as support and has high requirements for hardware equipment.

[0091] In view of the problems of low accuracy and low automation of existing mining road surface waterlogging detection, the present application provides a mining road surface waterlogging detection method.

[0092] Figure 1 Schematic diagram of the process of detecting waterlogging on mining roads provided by the embodiment of the present invention Figure 1 .like Figure 1 As shown, the method for detecting waterlogging on mining roads provided by the embodiment of the present invention includes the following steps:

[0093] S101. Acquire point cloud data in a first coordinate system.

[0094] Specifically, the point cloud data includes: coordinate information and reflection intensity of three-dimensional points of the open-pit mine road; the first coordinate system refers to a coordinate system with the laser radar as the coordinate origin;

[0095] The laser radar installed on the front side of the vehicle body scans the road information on both sides of the mine car's forward direction to obtain the open-pit mine road point cloud data. The point cloud data contains three-dimensional coordinate information xyz and reflection intensity. The positive direction of the laser radar coordinate system's x-axis is the vehicle's forward direction, the positive direction of the y-axis is the left side of the vehicle's forward direction, and the positive direction of the z-axis is vertically upward. The vehicle body coordinate system takes the center of the rear axle of the vehicle body as the coordinate origin, the positive direction of the x-axis is the vehicle's forward direction, the positive direction of the y-axis is the left side of the vehicle's forward direction, and the positive direction of the z-axis is vertically upward.

[0096] S102: Map the point cloud data to a second coordinate system, and construct a feature image of the point cloud data according to the coordinate information of each three-dimensional point.

[0097] Specifically, the second coordinate system refers to a coordinate system with the center of the rear axle of the vehicle body as the coordinate origin; the feature image is used to indicate the geometric features of the point cloud data in the second coordinate system;

[0098] Ensure that the LiDAR equipment is functioning properly and has been correctly installed in the appropriate location, such as on top of a vehicle or fixed at a specific observation point, so that the open-pit mine road can be effectively scanned. Aim the LiDAR at the open-pit mine road and turn on the LiDAR for scanning. The LiDAR emits a laser beam and receives the reflected laser signal;

[0099] For each laser beam emitted, the LiDAR measures its flight time and reflection intensity. The flight time can be used to calculate the distance between the object illuminated by the laser beam and the LiDAR. Combined with the angle information of the LiDAR, the coordinate information of the 3D point on the road in the open-pit mine area in the first coordinate system with the LiDAR as the coordinate origin can be determined;

[0100] At the same time, the reflection intensity information of each 3D point is recorded. The reflection intensity can reflect the material characteristics of the object surface and other information.

[0101] S103, obtaining the road boundary of the open-pit mine road from the feature image through a preset segmentation model.

[0102] Specifically, the segmentation model is trained through historical feature images;

[0103] According to the point cloud data, each pixel of each feature image is assigned a corresponding category label, a data set is established, and an appropriate deep learning semantic segmentation model is selected for training to obtain a feature image semantic segmentation model. The point cloud feature image is fed into the segmentation model to initially obtain the road boundary. Each road boundary is taken as a closed area, and the (inner_circle) algorithm is used to calculate the maximum inscribed circle of the closed area. The maximum inscribed circle point is obtained and the closed area is divided into two. The (inner_circle) algorithm is used to calculate the maximum inscribed circle for each of the two newly generated closed areas, and the maximum inscribed circle point is obtained and the closed area is further divided. This cycle is repeated until the maximum inscribed circle diameter of the closed area is less than the set threshold. The set of all maximum inscribed circle points is the road boundary.

[0104] S104. Determine a waterlogged area from a first closed area surrounded by a road boundary according to the coordinate information and reflection intensity of each three-dimensional point.

[0105] Specifically, the point cloud data is first filtered out according to the obtained road boundary, the index information of the point cloud within the boundary is retained, and the point cloud within the boundary in the original point cloud data is extracted using the index information;

[0106] Then, based on the characteristic that the reflection intensity of water is weak when the laser radar scans it, a certain threshold is set to filter out the point cloud with high reflection intensity, and preliminarily determine the area of ​​water to be detected;

[0107] The point cloud data of the area to be detected is mapped, and the (x, y) coordinate values ​​of the point cloud are mapped to the grid matrix. The point cloud data is divided based on the grid. For each grid, the height difference, average height and average reflection intensity of the point cloud cluster in the grid are calculated to obtain the feature matrix. The matrix is ​​converted into a grayscale image, and the edge recognition operator is used to perform edge recognition on the grayscale image. The recognition result is upgraded to obtain the road surface waterlogging detection result in a three-dimensional real scene.

[0108] The method for detecting water accumulation on mining roads in an embodiment of the present invention obtains point cloud data located in a first coordinate system; wherein the point cloud data includes: coordinate information and reflection intensity of three-dimensional points of the mining roads; the first coordinate system refers to a coordinate system with a laser radar as the coordinate origin; the point cloud data is mapped to a second coordinate system, and a feature image of the point cloud data is constructed based on the coordinate information of each three-dimensional point; wherein the second coordinate system refers to a coordinate system with the center of the rear axle of the vehicle body as the coordinate origin; the feature image is used to indicate the geometric features of the point cloud data in the second coordinate system; through a preset segmentation model, the road boundary of the mining road is obtained from the feature image; wherein the segmentation model is obtained by training with historical feature images; based on the coordinate information and reflection intensity of each three-dimensional point, a water accumulation area is determined from the first closed area surrounded by the road boundary. The following technical effects are achieved: raster conversion is performed by acquiring multi-frame laser point cloud information; each grid unit is processed according to the vehicle positioning information, and coordinate conversion is performed after processing; when the first water surface area is not greater than the preset first threshold, the corresponding contour is filtered out; according to the remaining contour, the position of the road surface water in the first coordinate system is determined; segmentation is performed according to the semantic segmentation annotation of the water accumulation to obtain the segmentation result, which is fused as the water accumulation recognition result, and the point cloud convex hull contour is recognized to obtain the water accumulation recognition result; the road boundary features of the three-dimensional point cloud data are extracted by a deep learning method, and the road boundary features are fitted using a road model-based method to obtain the road boundary; the reflectivity information of the point cloud is used to distinguish between the water accumulation area and the non-water accumulation area, so as to realize the detection of road surface water accumulation.

[0109] Figure 2 Schematic diagram of the process of detecting waterlogging on mining roads provided by the embodiment of the present invention Figure 2 .like Figure 2As shown, the method for detecting waterlogging on mining roads provided in this embodiment further describes the method for detecting waterlogging on mining roads, in which point cloud data is mapped to a second coordinate system, and a feature image of the point cloud data is constructed based on the coordinate information of each three-dimensional point. A grid matrix is ​​preset in the second coordinate system, and the grid matrix includes a plurality of first grids. The method for detecting waterlogging on mining roads provided in this embodiment includes the following steps:

[0110] S201. Acquire point cloud data in a first coordinate system.

[0111] The implementation of S201 is similar to that of S101 in the previous embodiment of the present invention, and will not be described in detail here.

[0112] S202, mapping the point cloud data to a grid matrix.

[0113] Specifically, the (x, y) coordinate values ​​of the point cloud are mapped to the grid matrix according to the coordinate transformation, and the point cloud data is divided based on the grid to form the grid matrix of the point cloud. The conversion formula is:

[0114]

[0115]

[0116] The grid matrix uses a Cartesian coordinate system with the lower left corner as the origin, where the positive direction of the column axis is the rightward extension direction of the matrix, and the positive direction of the row axis is the upward extension direction of the matrix; round represents the rounding operation, and the coordinate conversion result is rounded to ensure that the continuous floating-point coordinates (x, y) are converted to discrete integer grid coordinates (column, row); represents the minimum value of the x-axis direction in the point cloud data, represents the minimum value of the y-axis direction in the point cloud data, and a represents the length and width of the three-dimensional columnar area represented by each point in the grid matrix.

[0117] S203 . Divide the multiple three-dimensional points based on the multiple first grids to obtain a point cloud cluster of each first grid.

[0118] Specifically, first, the size of the first grid is determined according to the range of the open-pit mine road and the density of the point cloud data. For example, if the road range is large and the point cloud is sparse, a larger grid can be selected; conversely, if the road range is small and the point cloud is dense, a smaller grid can be selected. The coordinate system of the grid is determined to ensure that it corresponds to the first coordinate system with the laser radar as the coordinate origin, so as to accurately divide the three-dimensional points;

[0119] For each 3D point, determine the first grid it is in based on its coordinate information. The grid to which it belongs can be determined by comparing the coordinate value of the 3D point with the size and position of the grid through simple mathematical calculations;

[0120] The 3D points belonging to the same first grid are grouped into a point cloud cluster. A data structure such as a list or an array can be used to store the point cloud cluster of each grid.

[0121] S204: Obtain a grid feature matrix of the grid matrix according to the point cloud cluster of each first grid and the coordinate information of each three-dimensional point.

[0122] Specifically, organize the point cloud clusters of each first grid that has been divided, ensure that the three-dimensional point coordinate information in each point cloud cluster is accurate and has a clear correspondence with the corresponding first grid, and determine the type of grid features that need to be extracted, such as the average height of the point cloud cluster, the maximum height difference, the point cloud density, etc.;

[0123] According to the arrangement order of the first grid in the grid matrix, the feature vectors of each grid are arranged in sequence to form a two-dimensional grid feature matrix;

[0124] Assuming that the grid matrix has m rows and n columns, and the eigenvector of each grid has k eigenvalues, then the size of the grid feature matrix is ​​m×n×k.

[0125] S205, converting the grid feature matrix into a feature image of the point cloud data.

[0126] Specifically, the size of the feature image is determined according to the size and content of the grid feature matrix. For example, if the grid feature matrix has m rows and n columns, and the image pixel size corresponding to each grid is moderate, the width of the feature image can be set to n pixels and the height to m pixels;

[0127] For each eigenvalue in the raster feature matrix, design a mapping method to convert it into a color value. For example, linear mapping, logarithmic mapping, or other nonlinear mapping methods can be used to map the eigenvalue to a specific color interval based on its range. Suppose a certain eigenvalue represents the height of the point cloud, and the height range is from 0 to 100 meters. This range can be linearly mapped to a blue to red gradient in the color space, that is, a height of 0 corresponds to blue, a height of 100 meters corresponds to red, and intermediate height values ​​correspond to different degrees of blue and red mixed colors.

[0128] The technical effect of the scheme in this embodiment is: dividing the point cloud data in the vehicle body coordinate system into grids, each grid representing a columnar area in three-dimensional space, and representing the features in the grid point cloud data with set features for easy analysis, thereby improving the recognition accuracy of road surface detection and realizing quantitative analysis of road surface unevenness.

[0129] Figure 3 Schematic diagram of the process of detecting waterlogging on mining roads provided by the embodiment of the present invention Figure 3In order to describe the method for detecting waterlogging on mining roads in detail, the grid feature matrix of the grid matrix obtained according to the point cloud cluster of each first grid and the coordinate information of each three-dimensional point in the method for detecting waterlogging on mining roads is further described. Figure 3 As shown, the method for detecting waterlogging on mining roads provided by the embodiment of the present invention comprises the following steps:

[0130] S301 : obtaining the geometric features of each first grid according to the point cloud cluster of each first grid and the coordinate information of each three-dimensional point.

[0131] Specifically, the geometric features include: maximum height, average height, and height difference;

[0132] Average height: traverse the height coordinates of all 3D points in the grid point cloud cluster (assuming that the z-axis is the height direction in the coordinate system with the laser radar as the coordinate origin), add the height values ​​of all points, and then divide it by the total number of 3D points in the point cloud cluster to obtain the average height feature value of the grid;

[0133] Maximum height difference: Find the maximum height value and the minimum height value of the three-dimensional point in the grid point cloud cluster. The difference between the two is the maximum height difference eigenvalue.

[0134] S302: Fusing the geometric features of each first grid to obtain a grid feature matrix of the grid matrix.

[0135] Specifically, the geometric eigenvalues ​​of each grid are arranged in sequence to form a longer eigenvector. Then, these eigenvectors are combined according to the order of the grids in the matrix to form a grid feature matrix.

[0136] The technical effect of the solution in this embodiment is: by fusing the geometric features in the grid and using the fusion result as the grid feature, it is convenient to convert the grid feature into a feature image.

[0137] Figure 4 Schematic diagram of the process of detecting waterlogging on mining roads provided by the embodiment of the present invention Figure 4 In order to describe the method of detecting waterlogging on mining roads in detail, the boundary of mining roads is obtained from the feature image through the preset segmentation model. Figure 4 As shown, the method for detecting waterlogging on mining roads provided by the embodiment of the present invention comprises the following steps:

[0138] S401, inputting the feature image into a segmentation model to obtain an initial road boundary of an open-pit mine road output by the segmentation model.

[0139] Specifically, to collect and prepare training data, a large number of annotated open-pit mine road feature images are needed. The annotations should clearly indicate the boundaries of the roads. The model is trained using the prepared training data, and the parameters of the model are adjusted to improve the accuracy of segmentation.

[0140] The preprocessed feature image is input into the trained segmentation model. The model will classify the image pixel by pixel and assign each pixel to different categories, such as road, non-road, etc. The model outputs a segmentation result image with the same size as the input image, where the value of each pixel represents the category it belongs to. The segmentation result image is post-processed to extract the initial boundary of the road. Edge detection algorithms, such as edge detection, can be used to find the boundary between the road area and the non-road area in the image.

[0141] S402, performing cyclic segmentation on the second closed area enclosed by the initial boundary of the road, and obtaining the road boundary according to the segmentation result.

[0142] Specifically, starting from the second closed area enclosed by the initial boundary of the road, the first segmentation is performed, and the area is divided into multiple sub-areas according to a preset segmentation method, and the characteristic image information in the sub-area is checked to determine whether the area is closer to the characteristics of the road area. If a sub-area is considered to be more likely to be a road area, the area is further subdivided;

[0143] When the loop segmentation reaches the preset stop condition, the segmentation process stops;

[0144] Integrate all segmentation results to determine the final road boundary. For the sub-areas marked as road areas, merge and optimize their boundaries to form a continuous road boundary.

[0145] S403: Perform cyclic segmentation on the second closed area using an inscribed circle algorithm until the diameter of the largest inscribed circle obtained by segmentation is smaller than a preset diameter.

[0146] Specifically, each road boundary is taken as a closed area, and the inner_circle algorithm is used to calculate the maximum inscribed circle of the closed area. The point of the maximum inscribed circle is obtained and the closed area is divided into two. The inner_circle algorithm is used to calculate the maximum inscribed circle of the two newly generated closed areas, respectively, to obtain the maximum inscribed circle point and further divide the closed area. This cycle is repeated until the diameter of the maximum inscribed circle of the closed area is less than the set threshold.

[0147] S404: Obtain a road boundary according to the points of at least one maximum inscribed circle.

[0148] Specifically, the maximum inscribed circle point is obtained and the closed area is further divided, and this cycle is repeated until the maximum inscribed circle diameter of the closed area is less than the set threshold. The set of all maximum inscribed circle points is the road boundary.

[0149] The technical effect of the solution in this embodiment is: using a pre-set segmentation model and edge algorithm to calculate the road boundary, dividing the scanning area into road boundaries, and solving the problem of low accuracy in detecting waterlogging on mining roads in the prior art.

[0150] Figure 5 Schematic diagram of the process of detecting waterlogging on mining roads provided by the embodiment of the present invention Figure 5 In order to describe the method for detecting waterlogging on mining roads in detail, the waterlogging area is determined from the first closed area surrounded by the road boundary according to the coordinate information and reflection intensity of each three-dimensional point. Figure 5 As shown, the method for detecting waterlogging on mining roads provided by the embodiment of the present invention comprises the following steps:

[0151] S501. Filter out a plurality of target points from a plurality of three-dimensional points according to the reflection intensity of each three-dimensional point.

[0152] Specifically, traverse each 3D point:

[0153] For each 3D point, first determine whether it is located in the first closed area. This can be determined by comparing the coordinates of the 3D point with the boundary conditions of the first closed area. For example, if the first closed area is defined by a specific coordinate range, then determine whether the x, y, and z coordinates of the 3D point are within this range;

[0154] If the 3D point is in the first closed area, then check whether its reflection intensity is less than a preset intensity. Compare the reflection intensity of the point with the preset intensity;

[0155] If a 3D point satisfies both the conditions of being located in the first closed area and having a reflection intensity less than a preset intensity, it is marked as a target point;

[0156] A list or data structure can be used to store the filtered target points and record their coordinate information and reflection intensity.

[0157] S502: Determine an area to be detected from the first closed area according to multiple target points.

[0158] Specifically, the point cloud data is filtered out according to the obtained road boundary, the index information of the point cloud within the boundary is retained, and the point cloud within the boundary in the original point cloud data is extracted using the index information;

[0159] Based on the characteristic that the reflection intensity of accumulated water is weak when scanned by the laser radar, a certain threshold is set to filter out the point cloud with high reflection intensity, and preliminarily determine the area of ​​accumulated water to be detected.

[0160] S503 , rasterizing the area to be detected, dividing multiple target points based on multiple second grids in the area to be detected, and obtaining a point cloud cluster of each second grid.

[0161] Specifically, the area to be detected is divided into a plurality of second grids according to the determined grid size and shape. A regular grid grid can be formed by dividing the coordinate axis at a certain interval on the two-dimensional plane;

[0162] Traverse all target points, and for each target point, determine the second grid where it is located based on its coordinate information. The grid to which it belongs can be determined by comparing the coordinates of the target point with the boundary range of each grid, and the target point can be assigned to the corresponding second grid;

[0163] Forming a point cloud cluster, for each second grid, collect all target points assigned to the grid to form a point cloud cluster, and a data structure such as a list or an array can be used to store the point cloud cluster of each second grid.

[0164] S504: Determine a waterlogging area in the first closed area according to the point cloud cluster of each second grid, and the coordinate information and reflection intensity of each target point.

[0165] Specifically, the height difference, average height and average reflection intensity of the point cloud clusters within the grid are calculated to obtain three feature matrices, which are then converted into grayscale images;

[0166] The edge recognition operator is used to perform edge recognition on the grayscale image to obtain the edges of the waterlogged area in three feature dimensions. The intersection of the three edge results is taken to obtain the waterlogged area detection result in the two-dimensional feature map. Finally, the image result is upgraded to obtain the road waterlogging detection result in the three-dimensional real scene.

[0167] Optionally, in this embodiment, before mapping the point cloud data to the second coordinate system and constructing a feature image of the point cloud data according to the coordinate information of each three-dimensional point, the method further includes:

[0168] Preprocess the coordinate information of each 3D point.

[0169] Specifically, preprocessing includes denoising, correction of geometric errors and normalization.

[0170] Optionally, in this embodiment, according to the coordinate information and reflection intensity of each three-dimensional point, the method further includes determining the water accumulation area from the first closed area surrounded by the road boundary;

[0171] The two-dimensional feature image is upgraded to obtain the water accumulation area detection result in the three-dimensional scene.

[0172] Specifically, features related to the water accumulation area are further extracted from the two-dimensional feature image, and these features are analyzed to determine the possible water accumulation area;

[0173] The water accumulation area information in the two-dimensional image is associated with the three-dimensional scene data. The pixel position in the two-dimensional image can be matched with the point in the three-dimensional scene through methods such as coordinate transformation or feature matching.

[0174] The technical effect of the scheme in this embodiment is: extracting the point cloud within the road boundary and determining the waterlogging area to be detected based on the reflectivity intensity information, integrating the height difference, average height and average reflection intensity features to perform waterlogging edge recognition and detection, and upgrading the detection results to obtain road surface waterlogging detection results in real scenarios, which is used to solve the problems of low accuracy and low degree of automation in open-pit mining area road surface waterlogging detection in the prior art.

[0175] Figure 6 The schematic diagram of the structure of the mining road water surface detection device provided by the embodiment of the present invention. Figure 6 As shown, an embodiment of the present invention further provides a device for detecting water on a mining road, the device further comprising:

[0176] An acquisition module is used to acquire point cloud data located in a first coordinate system; wherein the point cloud data includes: coordinate information and reflection intensity of three-dimensional points of open-pit mine roads; the first coordinate system refers to a coordinate system with the laser radar as the coordinate origin;

[0177] A mapping module, used to map the point cloud data to a second coordinate system, and construct a feature image of the point cloud data according to the coordinate information of each three-dimensional point; wherein the second coordinate system refers to a coordinate system with the center of the rear axle of the vehicle body as the coordinate origin; and the feature image is used to indicate the geometric features of the point cloud data in the second coordinate system;

[0178] A segmentation module is used to obtain the road boundary of the open-pit mine road from the feature image through a preset segmentation model; wherein the segmentation model is obtained by training the historical feature image;

[0179] The determination module is used to determine the water accumulation area from the first closed area surrounded by the road boundary according to the coordinate information and reflection intensity of each three-dimensional point.

[0180] Optionally, the mapping module of the device as shown above includes:

[0181] A first mapping module, used for mapping point cloud data to a grid matrix;

[0182] A division module, used for dividing a plurality of three-dimensional points based on a plurality of first grids to obtain a point cloud cluster of each first grid;

[0183] A first calculation module, used for obtaining a grid feature matrix of the grid matrix according to the point cloud cluster of each first grid and the coordinate information of each three-dimensional point;

[0184] The conversion module is used to convert the raster feature matrix into a feature image of point cloud data.

[0185] Optionally, in the device shown above, the first computing module includes:

[0186] A second calculation module is used to obtain the geometric features of each first grid according to the point cloud cluster of each first grid and the coordinate information of each three-dimensional point; wherein the geometric features include: maximum height, average height and height difference;

[0187] The fusion module is used to fuse the geometric features of each first grid to obtain a grid feature matrix of the grid matrix.

[0188] Optionally, the device as shown above, the segmentation module, includes:

[0189] The first segmentation module is used to input the feature image into the segmentation model to obtain the initial boundary of the open-pit mine road output by the segmentation model;

[0190] The first cyclic segmentation module is used to perform cyclic segmentation on the second closed area enclosed by the initial boundary of the road, and obtain the road boundary according to the segmentation result.

[0191] Optionally, in the device shown above, the first cycle segmentation module includes:

[0192] A second cyclic segmentation module is used to perform cyclic segmentation on the second closed area by using an inscribed circle algorithm until the diameter of the largest inscribed circle obtained by segmentation is smaller than a preset diameter;

[0193] The collection module is used to obtain the road boundary according to the points of at least one maximum inscribed circle.

[0194] Optionally, the device as shown above, the determination module includes:

[0195] A screening module, used to screen out a plurality of target points from a plurality of three-dimensional points according to the reflection intensity of each three-dimensional point; wherein each target point is a three-dimensional point located in the first closed area, and the reflection intensity of each target point is less than a preset intensity;

[0196] A first determination module, used to determine the area to be detected from the first closed area according to the multiple target points;

[0197] A second division module is used to rasterize the area to be detected, divide multiple target points based on multiple second grids of the area to be detected, and obtain a point cloud cluster of each second grid;

[0198] The second determination module is used to determine the water accumulation area from the first closed area according to the point cloud cluster of each second grid and the coordinate information and reflection intensity of each target point.

[0199] Optionally, the mapping module of the device shown above further includes:

[0200] A preprocessing module is used to preprocess the coordinate information of each three-dimensional point; wherein the preprocessing includes denoising, correcting geometric errors and normalization;

[0201] The closed area is a two-dimensional feature image, and the determination module also includes:

[0202] The image processing module is used to upgrade the two-dimensional feature image to obtain the water accumulation area detection result in the three-dimensional scene.

[0203] Figure 7 Schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. Figure 7 As shown, an embodiment of the present invention further provides an electronic device, which includes: at least one processor 710 and a memory 720, and the electronic device also includes a communication component 730, wherein the processor 710, the memory 720 and the communication component 730 are connected via a bus 740.

[0204] In the specific implementation process, at least one processor 710 executes the computer execution instructions stored in the memory 720, so that the at least one processor 710 is used to implement the road surface water detection method for open-pit mine roads in the above embodiment.

[0205] The specific implementation process of the processor 710 can be found in the above method embodiment, and its implementation principle and technical effect are similar, so this embodiment will not be repeated here.

[0206] In the above embodiments, it should be understood that the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), etc. A general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in the invention may be directly implemented as being executed by a hardware processor, or may be executed by a combination of hardware and software modules in the processor.

[0207] The memory 720 may include a high-speed RAM memory, and may also include a non-volatile storage NVM, such as at least one disk storage.

[0208] The bus 740 may be an Industry Standard Architecture (Industry Standard Architecture), an ISA bus, a Peripheral Component (PCI) bus, or an Extended Industry Standard Architecture (Extended Industry Standard Architecture, EISA) bus, etc. The bus 740 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, the bus in the drawings of the present application is not limited to only one bus or one type of bus.

[0209] The above-mentioned functions implemented by the electronic device and the main control device introduce the scheme provided by the embodiment of the present invention. It can be understood that in order to implement the above-mentioned functions, the electronic device or the main control device includes a hardware structure and / or software module corresponding to the execution of each function. In combination with the units and algorithm steps of each example described in the embodiment disclosed in the embodiment of the present invention, the embodiment of the present invention can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the technical solution of the embodiment of the present invention.

[0210] The embodiment of the present application also provides a computer-readable storage medium, in which computer-executable instructions are stored, and when the computer-executable instructions are executed by a processor, they are used to implement the mining road waterlogging detection method of the above embodiment. In the specific implementation of the above mining road waterlogging detection method, each module can be implemented as a processor.

[0211] The above-mentioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk. The readable storage medium can be any available medium that can be accessed by a general or special-purpose computer.

[0212] An exemplary readable storage medium is coupled to a processor so that the processor can read information from the readable storage medium and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can be located in an application specific integrated circuit (ASIC). Of course, the processor and the readable storage medium can also exist as discrete components in an electronic device or a main control device.

[0213] An embodiment of the present application also provides a computer program product, including a computer program, which, when executed by a processor, is used to implement the method for detecting waterlogging on mining roads of the above embodiment.

[0214] The computer program is stored in a readable storage medium. At least one processor can read the computer program from the readable storage medium. At least one processor executes the computer program to execute the solution provided in any of the above embodiments.

[0215] Those skilled in the art can understand that all or part of the steps of the above method embodiments can be completed by hardware related to program instructions. The above program can be stored in a computer-readable storage medium. When the program is executed, the steps of the above method embodiments are executed; and the above storage medium includes: ROM, RAM, disk or optical disk and other media that can store program codes.

[0216] So far, the technical solution of the present application has been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present application is obviously not limited to these specific embodiments, and the above embodiments are only used to illustrate the technical solution of the present application rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for detecting waterlogging on mining roads, characterized in that: The method is applied to an intelligent driving vehicle, and the method includes: Acquire point cloud data in a first coordinate system; wherein the point cloud data includes: coordinate information and reflection intensity of three-dimensional points of an open-pit mine road; the first coordinate system refers to a coordinate system with a laser radar as a coordinate origin; Mapping the point cloud data to a second coordinate system, and constructing a feature image of the point cloud data according to the coordinate information of each of the three-dimensional points; wherein the second coordinate system refers to a coordinate system with the center of the rear axle of the vehicle body as the coordinate origin; and the feature image is used to indicate the geometric features of the point cloud data in the second coordinate system; Obtaining the road boundary of the open-pit mine road from the feature image through a preset segmentation model; wherein the segmentation model is obtained through training of historical feature images; A waterlogged area is determined from a first closed area surrounded by the road boundary according to the coordinate information and reflection intensity of each of the three-dimensional points.

2. The method according to claim 1, characterized in that A grid matrix is ​​preset in the second coordinate system, and the grid matrix includes a plurality of first grids; Mapping the point cloud data to a second coordinate system and constructing a feature image of the point cloud data according to the coordinate information of each of the three-dimensional points includes: Mapping the point cloud data to the grid matrix; Divide the plurality of three-dimensional points based on the plurality of first grids to obtain a point cloud cluster of each of the first grids; Obtaining a grid feature matrix of the grid matrix according to each point cloud cluster of the first grid and coordinate information of each three-dimensional point; The grid feature matrix is ​​converted into a feature image of the point cloud data.

3. The method according to claim 2, characterized in that The step of obtaining a grid feature matrix of the grid matrix according to each of the point cloud clusters of the first grid and the coordinate information of each of the three-dimensional points includes: According to the point cloud cluster of each first grid and the coordinate information of each three-dimensional point, the geometric features of each first grid are obtained; wherein the geometric features include: maximum height, average height and height difference; The geometric features of each of the first grids are fused to obtain a grid feature matrix of the grid matrix.

4. The method according to claim 1, characterized in that The step of obtaining the road boundary of the open-pit mine road from the feature image by using a preset segmentation model includes: Inputting the feature image into the segmentation model to obtain an initial road boundary of the open-pit mine road output by the segmentation model; The second closed area enclosed by the initial boundary of the road is segmented cyclically, and the road boundary is obtained according to the segmentation result.

5. The method according to claim 4, characterized in that The step of cyclically segmenting the second closed area enclosed by the initial boundary of the road and obtaining the boundary of the road according to the segmentation result includes: The second closed area is segmented cyclically by an inscribed circle algorithm until the diameter of the largest inscribed circle segmented is smaller than a preset diameter; The road boundary is obtained according to at least one point of the maximum inscribed circle.

6. The method according to claim 1, characterized in that The step of determining a waterlogged area from a first closed area surrounded by the road boundary according to the coordinate information and the reflection intensity of each of the three-dimensional points includes: According to the reflection intensity of each of the three-dimensional points, a plurality of target points are screened out from the plurality of the three-dimensional points; wherein each of the target points is a three-dimensional point located in the first closed area, and the reflection intensity of each of the target points is less than a preset intensity; Determine an area to be detected from the first closed area according to the plurality of target points; Rasterizing the area to be detected, dividing the plurality of target points based on the plurality of second grids of the area to be detected, and obtaining a point cloud cluster of each of the second grids; The water accumulation area is determined from the first closed area according to each point cloud cluster of the second grid, and the coordinate information and reflection intensity of each target point.

7. The method according to any one of claims 1 to 6, characterized in that: Before mapping the point cloud data to a second coordinate system and constructing a feature image of the point cloud data according to the coordinate information of each of the three-dimensional points, the method further includes: Preprocessing the coordinate information of each of the three-dimensional points; wherein the preprocessing includes denoising, correcting geometric errors and normalization; The closed area is a two-dimensional feature image, and the method further includes: determining the waterlogged area from the first closed area surrounded by the road boundary according to the coordinate information and reflection intensity of each of the three-dimensional points. The two-dimensional feature image is upgraded to obtain a water accumulation area detection result in a three-dimensional scene.

8. A device for detecting waterlogging on mining roads, characterized in that: The device is applied to an intelligent driving vehicle, and the device includes: An acquisition module is used to acquire point cloud data located in a first coordinate system; wherein the point cloud data includes: coordinate information and reflection intensity of three-dimensional points of open-pit mine roads; the first coordinate system refers to a coordinate system with the laser radar as the coordinate origin; A mapping module, used to map the point cloud data to a second coordinate system, and construct a feature image of the point cloud data according to the coordinate information of each of the three-dimensional points; wherein the second coordinate system refers to a coordinate system with the center of the rear axle of the vehicle body as the coordinate origin; and the feature image is used to indicate the geometric features of the point cloud data in the second coordinate system; A segmentation module, used to obtain the road boundary of the open-pit mine road from the feature image by using a preset segmentation model; wherein the segmentation model is obtained by training the historical feature image; The determination module is used to determine the water accumulation area from the first closed area surrounded by the road boundary according to the coordinate information and reflection intensity of each of the three-dimensional points.

9. An electronic device, characterized in that: include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executable instructions; When the processor executes the computer-executable instructions stored in the memory, it is used to implement the method for detecting waterlogging on mining roads as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method for detecting waterlogging on mining roads as described in any one of claims 1 to 7.

Images