A point cloud rendering method and system for unstructured roads in a mining area

Abstract

The application belongs to the technical field of unmanned driving, and specifically discloses a point cloud rendering method and system for unstructured roads in a mining area, which comprises the following steps: obtaining trajectory points of a vehicle and point clouds of a road in a mining area scanned by a laser radar on the vehicle; the trajectory points are selected at equal intervals to obtain trajectory center points; at the same time, the relative elevation values of the point clouds relative to the trajectory center points are calculated, and the relative elevation values are written into the attributes of the point clouds; a mining area map is extracted according to the point clouds; the point clouds are rendered according to the attributes, and the mining area map is classified and rendered according to different types; and the rendered mining area map and point clouds are superimposed; the method has the following advantages: the rendered mining area map and point clouds are superimposed and displayed together, the accuracy of the extracted data can be judged according to the superimposed result, and accurate point cloud data is provided for subsequent high-precision map data editing.

Description

Technical Field

[0001] This invention relates to the field of autonomous driving technology, and more specifically, to a point cloud rendering method and system for unstructured roads in mining areas. Background Technology

[0002] Autonomous driving refers to intelligent unmanned driving control technology in which vehicles perceive the road environment through onboard sensor systems, rely on computer systems to plan driving routes and reach their destinations. It uses onboard sensors to perceive the surrounding environment of the vehicle and controls the vehicle's steering and speed based on the perceived road, vehicle position, and obstacle information, thereby enabling the vehicle to drive safely and reliably on the road. It integrates many technologies such as automatic control, system architecture, artificial intelligence, and visual computing. As a type of autonomous driving, unmanned driving in mining areas is applied to specific sites and specific vehicles in mining areas. The mining areas specifically refer to open-pit mines, such as iron mines and coal mines, and the vehicles used are large mining trucks and wide-body vehicles.

[0003] LiDAR scanning systems can quickly and accurately acquire high-precision 3D spatial information. Extracting high-precision 3D road information from LiDAR point clouds plays a crucial role in fields such as 3D navigation and autonomous driving. However, vehicle-mounted LiDAR point clouds suffer from drawbacks such as uneven density distribution, large data volume, incomplete data, and complex scenes, which affect the completeness and accuracy of road extraction. Point cloud rendering can be used to classify and render points based on their different attribute values. For example, using the elevation attribute of the point cloud allows viewing its elevation distribution; RGB rendering can also be used to view the geographic information displayed by the point cloud in true color.

[0004] Existing point cloud classification rendering methods, when using elevation rendering for relatively flat roads or work areas, cannot distinguish between the road surface and road boundaries. Using RGB rendering also presents difficulties in differentiating the road surface and boundaries in the monotonous colors of open-pit mines. For example, when there is light and shadow occlusion, the road boundary cannot be distinguished from the road surface when rendered using RGB rendering.

[0005] To address these issues, a point cloud rendering method and system for unstructured roads in mining areas is proposed. Summary of the Invention

[0006] The present invention aims to provide a point cloud rendering method and system for unstructured roads in mining areas, so as to solve or improve the problem of identifying road boundaries in the point cloud classification rendering of open-pit mines.

[0007] In view of this, the first aspect of the present invention is to provide a point cloud rendering method for unstructured roads in mining areas.

[0008] A second aspect of the present invention is to provide a point cloud rendering system for unstructured roads in mining areas.

[0009] The first aspect of the present invention provides a point cloud rendering method for unstructured roads in a mining area, comprising the following steps: S1, acquiring the trajectory points of a vehicle and the point cloud of the mining area road scanned by a lidar on the vehicle; S2, selecting the trajectory points at equal intervals to obtain the trajectory center point, calculating the slope of the current road based on the position of the trajectory center point, and determining whether the current road is a slope based on the slope; if so, reducing the interval and reselecting the trajectory center point; S3, calculating the relative elevation value of the point cloud relative to the trajectory center point in S3 at the same time, and writing the relative elevation value into the attributes of the point cloud; S4, obtaining the position on a known map based on the latitude and longitude of the trajectory center point, and determining the feature type of the map where the point cloud is located; S5, rendering the point cloud according to the attributes, and rendering the map with color according to the feature type, performing coordinate projection transformation on the rendered map, and overlaying it with the point cloud.

[0010] This invention provides a point cloud rendering method for unstructured roads in mining areas. By overlaying the rendered mining area map and point cloud together for display, the accuracy of the extracted data can be determined based on the overlay result. Based on the difference between the overlaid map and the real-time point cloud, it can be determined whether the current high-precision map needs to be updated. At the same time, because the point cloud has already been rendered, the difference between the map and the point cloud can be clearly seen after it is displayed on the map, providing accurate point cloud data for subsequent high-precision map data editing, and enabling precise manual editing and updates.

[0011] By using the relative elevation value of the point cloud with respect to the center point of the trajectory as an attribute of the point cloud data, the actual geographic information represented by the point cloud data of the open-pit mine can be clearly identified;

[0012] The calculated relative height can be used to determine the boundary of the road point cloud, providing a basis for verifying the accuracy of the boundary data extracted from the point cloud. Due to the relative height, the ground points and retaining wall points have a clear color difference, and the boundary can be clearly seen.

[0013] Finally, the rendered results are displayed, providing accurate data updates for high-precision mine map data. This enables the creation of more accurate and error-free high-precision map data, providing safe and precise data assurance for unmanned mining operations.

[0014] In addition, the technical solutions provided by embodiments of the present invention may also have the following additional technical features:

[0015] In any of the above technical solutions, the vehicle outputs 20 trajectory points per second, and step S2 specifically includes: S21, filtering the trajectory points output per second at 2m intervals and calculating the slope between adjacent trajectory points; S22, determining whether the road where the current trajectory point is located is inclined based on the slope, if so, proceeding to S23 and clearing the filtering results, otherwise proceeding to S24; S23, filtering the trajectory points at the current moment at 1m intervals; S24, outputting the currently filtered trajectory point as the trajectory center point.

[0016] In this technical solution, based on the number of trajectory points actually output per second by the LiDAR on the vehicle, the corresponding trajectory points on the road at the same time are divided and filtered at equal distances to ensure that the horizontal distance between the selected adjacent trajectory points is similar and to avoid excessive errors in individual data.

[0017] By checking whether the current road is tilted, the corresponding trajectory points of the tilted road need to be re-filtered, and a smaller filtering gap needs to be selected to reduce the elevation difference between adjacent trajectory points, so that they can be in a reasonable color value range in subsequent rendering and ensure the normal output of the rendering results.

[0018] Before writing relative elevation values ​​into the point cloud data, the slope between adjacent trajectory points is used to determine whether the road is tilted. Different interval sizes are used for filtering to adapt to different roads and ensure the stability of the detection results for roads in the mining area.

[0019] In any of the above technical solutions, the following rule is used to determine whether the slope is tilted in S22: the slope is compared with 5°. If the slope is greater than 5°, the road where the trajectory point is located is determined to be tilted; if the slope is less than 5°, the road where the trajectory point is located is determined to be non-tilted.

[0020] In this technical solution, a value of 5° is set to judge the slope. Roads with a slope greater than 5° are considered inclined roads. By setting the slope judgment value in a targeted manner, the final judgment standard for whether a road is inclined can be more adapted to the actual situation of roads in mining areas.

[0021] In any of the above technical solutions, the data of the trajectory points and point clouds respectively include: latitude and longitude and elevation.

[0022] In this technical solution, the data recording trajectory points and point clouds include latitude, longitude, and elevation, which enables a better understanding of the road conditions in the mining area and ensures the final rendering accuracy.

[0023] In any of the above technical solutions, step S3 specifically includes: S31, calculating the relative elevation value of the point cloud relative to the elevation of the trajectory center point based on the elevation of the trajectory center point at the same time; S32, stretching and converting the relative elevation value and writing it into the userdata attribute of the point cloud.

[0024] In this technical solution, the elevations of the selected trajectory center point and the point cloud are calculated simultaneously to obtain a relative elevation value. This avoids calculation errors caused by different surveying locations due to time errors, ensuring the actual correspondence between the trajectory center point and the point cloud, reducing misalignment errors, and writing the relative elevation value into the point cloud's attributes to correspond with the subsequent rendering color value. This allows the actual height to be reflected in the subsequent rendering color, avoiding the difficulty of distinguishing road surfaces and boundaries in the monotonous color of an open-pit mine using RGB rendering.

[0025] Specifically, the relative elevation value of the point cloud with respect to the center point of the trajectory is calculated using the following formula:

[0026] Δh=h 点云 -h 轨迹中心点 ;

[0027] Where Δh is the relative elevation value, h 点云 For the elevation of the point cloud, h 轨迹中心点 The elevation of the center point of the trajectory.

[0028] In any of the above technical solutions, the stretching conversion of the relative elevation value is calculated using the following formula: x1=(x2+0.5)*100 where x1 is the relative elevation value of the point cloud after stretching conversion and x2 is the relative elevation value of the point cloud before stretching conversion.

[0029] In this technical solution, a formula is used to calculate the relative elevation value and the color value of subsequent rendering, which ensures a stable correspondence between different values ​​and reduces the error of the rendered image. At the same time, the point cloud with different relative elevation values ​​can be pre-screened for the specific range of color values ​​to be rendered, ensuring that the color value range required for rendering will not be exceeded during subsequent rendering.

[0030] In any of the above technical solutions, before stretching and transforming the relative elevation value, step S32 further includes: filtering the point cloud based on the relative elevation value; wherein the filtering range is [-0.5, 2].

[0031] In this technical solution, since the color values ​​will have specific range requirements in the subsequent rendering, the point cloud with the relative elevation value attribute needs to be filtered within a certain range before rendering so that the point cloud can be completely processed by the subsequent rendering steps. The specific filtering range is [-0.5, 2].

[0032] In any of the above technical solutions, the mining area map includes: data of road surfaces and their corresponding boundaries, and data of work area surfaces and their corresponding boundaries; wherein, the work area is the site where mining operations take place.

[0033] In this technical solution, the data extraction of the mining area map specifically needs to include the data of the road surface and its corresponding boundary of the mining area, as well as the data of the working area and its corresponding boundary. While processing data as much as possible, it ensures that the necessary parts of the final road analysis are available for the mining truck to call up data under unmanned driving.

[0034] In any of the above technical solutions, the rendering in S5 uses the WebGL rendering engine, and the point cloud is a LAS point cloud.

[0035] A second aspect of the present invention provides a point cloud rendering system for unstructured roads in mining areas, comprising: a data acquisition module for acquiring point cloud data from a vehicle-mounted lidar and latitude and longitude location data; a key point selection module for selecting the center point of a trajectory; wherein the sampling interval is 2 meters, and if the current road is in an inclined state, the sampling interval is reduced for denser sampling; a feature calculation module for calculating the relative elevation value of the point cloud and storing it as a point cloud attribute feature; wherein the relative elevation value is stretched to a certain extent to distinguish relative height; a map projection module for unifying the real-time frame point cloud and the known map into the same coordinate system, and determining the working segment where the current point cloud is located based on the latitude and longitude data; and a rendering module for point cloud rendering, classifying and rendering the current point cloud according to the point cloud spatial attributes acquired by the map projection module and the point cloud relative elevation attributes acquired by the feature calculation module, and displaying it on a high-precision map to determine whether the map needs to be updated; wherein the system is used to implement the method as described in any one of the first aspects.

[0036] This invention provides a point cloud rendering system for unstructured roads in mining areas, which implements the steps of a point cloud rendering method for unstructured roads in mining areas as described in any of the above-mentioned technical solutions. Therefore, the point cloud rendering system for unstructured roads in mining areas proposed in this technical solution possesses all the beneficial effects of a point cloud rendering method for unstructured roads in mining areas as described in any of the above-mentioned technical solutions, and will not be elaborated further here.

[0037] The beneficial effects of this invention compared to the prior art are as follows:

[0038] By calculating relative elevation and stretching it, the height differences in geographic information can be amplified, and the actual geographic information represented by the point cloud data of open-pit mines can be clearly identified. At the same time, the road area and the retaining wall area have obvious color distinctions, which can clearly determine the boundary position and provide a basis for accuracy verification of the boundary data extracted from the point cloud.

[0039] Furthermore, by correlating with high-precision maps and comparing the differences between real-time point clouds and high-precision maps, it can provide accurate data updates for high-precision mine map data.

[0040] Additional aspects and advantages of embodiments of the invention will become apparent in the following description or may be learned by practice of embodiments of the invention. Attached Figure Description

[0041] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which:

[0042] Figure 1 This is a flowchart of the method of the present invention;

[0043] Figure 2 This is a schematic diagram of the flat-path acquisition frequency of the present invention;

[0044] Figure 3 This is a schematic diagram of the slope acquisition frequency of the present invention;

[0045] Figure 4 This is a flowchart of the relative elevation difference calculation method of the present invention;

[0046] Figure 5 This is a rendering result using traditional methods;

[0047] Figure 6 This is a rendering of the relative elevation effect of the present invention;

[0048] Figure 7 This is a system logic block diagram of the present invention. Detailed Implementation

[0049] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0050] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0051] Please see Figure 1-7 The following describes a point cloud rendering method and system for unstructured roads in mining areas, based on some embodiments of the present invention.

[0052] An embodiment of the first aspect of the present invention proposes a point cloud rendering method for unstructured roads in mining areas. In some embodiments of the present invention, such as... Figure 1-6 As shown, a point cloud rendering method for unstructured roads in a mining area is provided. This method includes:

[0053] The first step is to generate point cloud data based on the scanning data from the vehicle-mounted LiDAR:

[0054] After acquiring the LAS point cloud data of the mining area roads, it is also necessary to acquire the trajectory data of the acquisition vehicle. The data must include latitude, longitude and actual elevation.

[0055] The role of the vehicle-mounted lidar is to collect point cloud data from the mine.

[0056] The lidar of the point cloud acquisition vehicle normally outputs 20 trajectory points per second for left and right navigation. First, trajectory points are selected based on an interval of 2m. The slope is calculated based on the two selected trajectory points. If the angle is greater than 5°, it is defined as a slope; if it is less than 5°, it is defined as a flat road. If it is a slope, trajectory points are reselected at an interval of 1m.

[0057] The second step is to process the elevation difference between the scanned point cloud data and the elevation of the trajectory points at the same time, and write the result into the attributes of the point cloud data:

[0058] At the same point in time, calculate the relative elevation of the mine road point cloud data collected at the current time point relative to the elevation of the trajectory point. Crop the point cloud data whose relative elevation values ​​fall within the range of [-0.5, 2].

[0059] For point cloud data within a given range, the relative elevation values ​​are converted into color values ​​[0, 255] required for WebGL rendering using a formula.

[0060] x1 = (x2 + 0.5) * 100

[0061] Where x1 is the relative elevation value of the point cloud after stretching transformation, and x2 is the relative elevation value of the point cloud before stretching transformation.

[0062] Write the converted relative elevation data into the userdata property of the point cloud.

[0063] The third step is to extract high-precision map elements from the point cloud:

[0064] High-precision map data is extracted based on point cloud data. The main feature types of the extracted high-precision map data include: loading area, unloading area, parking area, and temporary parking area.

[0065] The fourth step involves the WebGL rendering engine loading and rendering the point cloud data. Then, the extracted high-precision map data is transformed into the same coordinate system as the point cloud through coordinate projection. Finally, it is rendered according to different map element types and overlaid onto the point cloud.

[0066] The WebGL rendering engine classifies and renders the processed point cloud data, while simultaneously loading extracted high-precision map data, and displays the two overlaid.

[0067] Finally, the WebGL rendering engine is used to load the point cloud data, and rendering is performed based on the userdata attribute of the point cloud. Then, the extracted high-precision map data is loaded and overlaid for display. Based on the overlay result, the accuracy of the extracted data is determined, and precise point cloud data is provided as a basis for subsequent high-precision map data editing, enabling precise manual editing and updates.

[0068] This invention provides a point cloud rendering method for unstructured roads in mining areas. For point cloud data of mining roads and work areas, rendering is performed based on the elevation difference between the point cloud and the lane centerline of the acquiring vehicle to distinguish the road surface and road boundaries. This method is used to verify the accuracy of road boundary data extracted from vehicle-mounted LiDAR. Specifically, it aims to verify the accuracy of point cloud road extraction, provide a basis for updating boundary data, and provide accurate and efficient base map data for unmanned driving in mining areas.

[0069] A second aspect of the present invention provides a point cloud rendering system for unstructured roads in mining areas. In some embodiments of the present invention, such as... Figure 7 As shown, a point cloud rendering system for unstructured roads in a mining area is provided. This system includes:

[0070] The data acquisition module is used to collect point cloud data and latitude and longitude location data from the vehicle-mounted lidar.

[0071] The key point selection module is used to select key trajectory points; the sampling interval is 2 meters. If the current road is inclined, the sampling interval is reduced and the sampling is more frequent.

[0072] The feature calculation module is used to calculate the relative elevation values ​​of the point cloud and save them as point cloud attribute features; the relative elevation values ​​are stretched to a certain extent to distinguish relative heights.

[0073] The map projection module is used to unify real-time frame point clouds and high-precision maps into the same coordinate system, and determine the current working area of ​​the point cloud based on latitude and longitude data.

[0074] The rendering module is used for point cloud rendering. Based on the spatial attributes of the point cloud obtained by the map projection module and the relative elevation attributes of the point cloud obtained by the feature calculation module, the current point cloud is classified and rendered, and displayed on the high-precision map to determine whether the map needs to be updated.

[0075] The system is used to implement the method as described in any one of the first aspects.

[0076] This invention provides a point cloud rendering system for unstructured roads in mining areas, which implements the steps of a point cloud rendering method for unstructured roads in mining areas as described in any of the above embodiments. Therefore, the point cloud rendering system for unstructured roads in mining areas proposed in this embodiment possesses all the beneficial effects of the point cloud rendering method for unstructured roads in mining areas as described in any of the above embodiments, and will not be elaborated further here.

[0077] In the description of this invention, it should be understood that the terms "longitudinal", "lateral", "up", "down", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this invention, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this invention.

[0078] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made by those skilled in the art to the technical solutions of the present invention without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A point cloud rendering method for unstructured roads in a mining area, characterized in that, Includes the following steps: S1, acquire the vehicle's trajectory points and the point cloud of the mining area road scanned by the vehicle's lidar; S2, Select the trajectory points at equal intervals to obtain the trajectory center point; The slope of the current road is calculated based on the position of the center point of the trajectory, and it is determined whether the current road is a slope based on the slope. If so, the interval is reduced and the center point of the trajectory is reselected. S3, at the same time, calculate the relative elevation value of the point cloud relative to the trajectory center point in S2, and write the relative elevation value into the attribute of the point cloud; S4. Obtain the location on a known map based on the latitude and longitude of the trajectory center point, and determine the feature type of the map where the point cloud is located; S5, render the point cloud according to the attributes, and render the map according to the type of the feature. Then, perform coordinate projection transformation on the rendered map and overlay it with the point cloud. The steps in S3 specifically include: S31, Calculate the relative elevation value of the point cloud relative to the elevation of the trajectory center point based on the elevation of the trajectory center point at the same time. S32, stretch and transform the relative elevation value and write it into the attribute of the point cloud; The stretching conversion of the relative elevation value is calculated using the following formula: wherein, is the relative elevation value of the point cloud after the stretch conversion, is the relative elevation value of the point cloud before the stretch conversion; In S5, the WebGL rendering engine is used to load point cloud data and render it according to the userdata attribute of the point cloud.

2. The method of rendering a point cloud of an unstructured mine road according to claim 1, wherein, The vehicle outputs 20 trajectory points per second, and the steps in S2 specifically include: S21, the trajectory points output per second are filtered at 2m intervals, and the slope between adjacent trajectory points after filtering is calculated; S22, determine whether the road where the current trajectory point is located is inclined by the slope. If so, proceed to S23 and clear the filtered results. Otherwise, proceed to S24. S23, filter the trajectory points at the current moment at 1m intervals; S24, output the currently selected trajectory points as the trajectory center points.

3. The method of rendering a point cloud of an unstructured mine road according to claim 2, wherein, Whether or not the structure in S22 is tilted is determined by the following rule: The slope is compared with 5°. If the slope is greater than 5°, the road where the trajectory point is located is determined to be inclined. If the slope is less than 5°, the road where the trajectory point is located is determined to be non-inclined.

4. The point cloud rendering method for unstructured roads in a mining area according to claim 1, characterized in that, The trajectory points and point cloud data include latitude, longitude, and elevation, respectively.

5. A point cloud rendering method for unstructured roads in a mining area according to claim 4, characterized in that, Before performing the stretching transformation on the relative elevation value, step S32 further includes: The point cloud is filtered based on the relative elevation values; The filtering range is [-0.5, 2].

6. A point cloud rendering method for unstructured roads in a mining area according to claim 1, characterized in that, The mining area map includes: data on road surfaces and their corresponding boundaries, and data on work area surfaces and their corresponding boundaries; The work area refers to the site where mining operations take place.

7. A point cloud rendering method for unstructured roads in a mining area according to claim 1, characterized in that, The rendering in S5 uses the WebGL rendering engine, and the point cloud is a LAS point cloud.

8. A point cloud rendering system for unstructured roads in mining areas, characterized in that, include: The data acquisition module is used to collect point cloud data and latitude and longitude location data from the vehicle-mounted lidar. The key point selection module is used to select the center point of the trajectory; the sampling interval is 2 meters. If the current road is inclined, the sampling interval is reduced and the sampling is more frequent. The feature calculation module is used to calculate the relative elevation values ​​of the point cloud and save them as point cloud attribute features; the relative elevation values ​​are stretched to a certain extent to distinguish relative heights. The map projection module is used to unify the real-time frame point cloud and the known map into the same coordinate system, and to determine the current working area of ​​the point cloud based on latitude and longitude data. The rendering module is used for point cloud rendering. Based on the spatial attributes of the point cloud obtained by the map projection module and the relative elevation attributes of the point cloud obtained by the feature calculation module, the current point cloud is classified and rendered, and displayed on the high-precision map to determine whether the map needs to be updated. The system is used to implement the method as described in any one of claims 1-7.

Images