Method, apparatus for superimposing laser point clouds and high-precision map and electronic device
The present application discloses a method, an apparatus for superimposing laser point clouds and a high-precision map and an electronic device, comprising: when superimposing the laser point clouds and the high-precision map for visualization, first performing dilution processing on non-road laser point clouds in the laser point clouds to be fused to obtain target non-road laser point clouds, which effectively reduces the amount of data of the laser point clouds to be fused; and performing segmentation processing on the map to be fused to obtain multi-level map data corresponding to the map to be fused, which effectively reduces the amount of data of the map to be fused. In this way, when superimposing the laser point clouds and the high-precision map for visualization, both non-road information may be visualized through the laser point clouds, and road information may be visualized through the high-precision map.
1. A method for superimposing laser point clouds and a high-precision map, comprising:
acquiring laser point clouds to be fused and a map to be fused respectively, wherein the map to be fused at least comprises a partial high-precision map;
performing dilution processing on non-road laser point clouds in the laser point clouds to be fused to obtain target non-road laser point clouds;
performing segmentation processing on the map to be fused to obtain multi-level map data corresponding to the map to be fused; and
performing superposition processing on the target non-road laser point clouds and the multi-level map data, and visualizing a superposition result;
wherein the performing dilution processing on non-road laser point clouds in the laser point clouds to be fused to obtain target non-road laser point clouds comprises:
determining a dilution ratio;
performing a weighted octree segmentation on the non-road laser point clouds according to the dilution ratio to obtain a non-road laser point cloud corresponding to each of multiple leaf nodes; and
calculating, for each leaf node, a center of gravity within a voxel of the non-road laser point cloud corresponding to the leaf node, and weighting an attribute value of the non-road laser point cloud corresponding to the leaf node to obtain an attribute value of the center of gravity; wherein the attribute value of the center of gravity is target non-road laser point cloud corresponding to the leaf node.
2. The method according to claim 1 , wherein the performing superposition processing on the target non-road laser point clouds and the multi-level map data comprises:
performing chunking processing on the target non-road laser point clouds to obtain multiple target point cloud data blocks corresponding to the target non-road laser point clouds; and
performing superposition processing on the multiple target point cloud data blocks and the multi-level map data based on a visualized view-frustum area range.
3. The method according to claim 2 , wherein the performing superposition processing on the multiple target point cloud data blocks and the multi-level map data based on a visualized view-frustum area range comprises:
screening out point cloud data blocks and map data that meet a multi-level of detail segmentation rule from the multiple target point cloud data blocks and the multi-level map data respectively based on the visualized view-frustum area range; and
performing superposition processing on the point cloud data blocks and the map data that meet the multi-level of detail segmentation rule.
4. The method according to claim 1 , wherein the method further comprises:
performing screening processing on the laser point clouds to be fused to obtain initial non-road laser point clouds; and
performing migration processing on elevations of the initial non-road laser point clouds to obtain the non-road laser point clouds.
5. The method according to claim 4 , wherein when the laser point clouds to be fused further comprises initial road laser point clouds, then the performing migration processing on elevations of the initial non-road laser point clouds to obtain the non-road laser point clouds comprises:
determining an average road elevation corresponding to single frame data in the laser point clouds to be fused;
subtracting the average road elevation from elevations of all laser points in the single frame of data to obtain single frame data after the migration processing; wherein the single frame data after the migration processing comprises road laser point clouds after performing the migration processing on the initial road laser point cloud; and
removing the road laser point clouds after the migration processing from the single frame data after the migration processing to obtain the non-road laser point clouds.
6. The method according to claim 4 , wherein when the laser point clouds to be fused do not comprise initial road laser point clouds, then the performing migration processing on elevations of the initial non-road laser point clouds to obtain the non-road laser point clouds comprises:
determining a previous frame of single frame data having been migrated that is closest to single frame data in the laser point clouds to be fused and a subsequent frame of single frame data having been migrated that is closest to the single frame data in the laser point clouds to be fused;
performing interpolation processing on the previous frame of single frame data and the subsequent frame of single frame data to obtain the single frame data after the migration processing; wherein, the single frame data after the migration processing comprises road laser point clouds after performing the migration processing on the initial road laser point cloud; and
removing the road laser point clouds after the migration processing from the single frame data after the migration processing to obtain the non-road laser point clouds.
7. The method according to claim 1 , wherein the acquiring laser point clouds to be fused comprises:
acquiring initial laser point clouds; and
performing optimizing processing on precision of the initial laser point clouds to obtain the laser point clouds to be fused; wherein precision of the laser point clouds to be fused is higher than the precision of the initial laser point clouds.
8. The method according to claim 1 , wherein when the map to be fused further comprises a standard precision map, the acquiring a map to be fused comprises:
performing optimizing processing on precision of the standard precision map to obtain a processed map; and
performing fusion on the processed map and the high-precision map to obtain the map to be fused.
9. A non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause a computer execute the method for superimposing the laser point clouds and the high-precision map according to claim 1 .
10. An apparatus for superimposing laser point clouds and a high-precision map, comprising:
at least one processor; and
a memory communicatively connected with the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor, to enable the at least one processor to:
acquire laser point clouds to be fused and a map to be fused respectively, wherein the map to be fused at least comprises a partial high-precision map;
perform dilution processing on non-road laser point clouds in the laser point clouds to be fused to obtain target non-road laser point clouds, and perform segmentation processing on the map to be fused to obtain multi-level map data corresponding to the map to be fused;
perform superposition processing on the target non-road laser point clouds and the multi-level map data; and
visualize a superposition result;
wherein the at least one processor is further enabled to:
determine a dilution ratio;
perform a weighted octree segmentation on the non-road laser point clouds according to the dilution ratio to obtain a non-road laser point cloud corresponding to each of the multiple leaf nodes; and
calculate, for each leaf node, a center of gravity within a voxel of the non-road laser point cloud corresponding to the leaf node, and perform weighting on an attribute value of the non-road laser point cloud corresponding to the leaf node to obtain an attribute value of the center of gravity; wherein the attribute value of the center of gravity is target non-road laser point cloud corresponding to the leaf node.
11. The apparatus according to claim 10 , wherein the at least one processor is further enabled to:
perform chunking processing on the target non-road laser point clouds to obtain multiple target point cloud data blocks corresponding to the target non-road laser point clouds; and
perform superposition processing on the multiple target point cloud data blocks and the multi-level map data based on a visualized view-frustum area range.
12. The apparatus according to claim 11 , wherein the at least one processor is further enabled to:
screen out point cloud data blocks and map data that meet a multi-level of detail segmentation rule from the multiple target point cloud data blocks and the multi-level map data respectively based on the visualized view-frustum area range;
perform superposition processing on the point cloud data blocks and the map data that meet the multi-level of detail segmentation rule.
13. The apparatus according to claim 10 , wherein the at least one processor is further enabled to:
perform screening processing on the laser point clouds to be fused to obtain initial non-road laser point clouds; and
perform migration processing on elevations of the initial non-road laser point clouds to obtain the non-road laser point clouds.
14. The apparatus according to claim 13 , wherein when the laser point clouds to be fused further comprises initial road laser point clouds, then the at least one processor is further enabled to:
determine an average road elevation corresponding to single frame data in the laser point clouds to be fused;
subtract the average road elevation from elevations of all laser points in the single frame data to obtain the single frame data after the migration processing; wherein the single frame data after the migration processing comprises the road laser point clouds after performing the migration processing on the initial road laser point clouds; and
remove the road laser point clouds after the migration processing from the single frame data after the migration processing to obtain the non-road laser point clouds.
15. The apparatus according to claim 13 , wherein when the laser point clouds to be fused do not comprise initial road laser point clouds, then the at least one processor is further enabled to:
determine a previous frame of single frame data having been migrated that is closest to single frame data in the laser point clouds to be fused and a subsequent frame of single frame data having been migrated that is closest to the single frame data in the laser point clouds to be fused;
perform interpolation processing on the previous frame of single frame data and the subsequent frame of single frame data to obtain the single frame data after the migration processing; wherein, the single frame data after the migration processing comprises the road laser point clouds after performing the migration processing on the initial road laser point clouds; and
remove the road laser point clouds after the migration processing from the single frame data after the migration processing to obtain the non-road laser point clouds.
16. The apparatus according to claim 10 , wherein the at least one processor is further enabled to:
acquire initial laser point clouds; and
perform optimizing processing on precision of the initial laser point clouds to obtain the laser point clouds to be fused; wherein the precision of the laser point clouds to be fused is higher than the precision of the initial laser point clouds.
17. The apparatus according to claim 10 , wherein when the map to be fused further comprises a standard precision map, the at least one processor is further enabled to:
perform optimizing processing on precision of the standard precision map to obtain a processed map; and
perform fusion on the processed map and the high-precision map to obtain the map to be fused.