METHOD AND DEVICE FOR FILTERING OUT NON-GROUND POINTS FROM POINT CLOUD, AND STORAGE MEDIUM
Methods and devices for filtering out non-ground points from a point cloud. A method includes acquiring a point cloud by scanning ground with light detection and ranging (LiDAR); calculating radial distances of adjacent points in the point cloud; determining key points based on a variation of the radial distances of adjacent points; and filtering out non-ground points from the point cloud based on the key points. Key points are determined based on the radial distances of adjacent points in the point cloud, and then the non-ground points are filtered out.
1 . A method for filtering out non-ground points from a point cloud, comprising:
acquiring a point cloud by scanning ground with light detection and ranging (LiDAR);
calculating radial distances of adjacent points in the point cloud;
determining key points based on a variation of the radial distances of adjacent points; and
filtering out non-ground points from the point cloud based on the key points.
2 . The method of claim 1 , wherein the adjacent points comprise points obtained by scanning the ground using adjacent laser heads of the LiDAR or points obtained by scanning the ground using one laser head of the LiDAR at adjacent scanning angles, and wherein the radial distance is a distance from a projection position of a point of the point cloud to a coordinate origin in the top view plane, and the coordinate origin in the top view plane is a projection of a laser origin in the top view plane.
3 . The method of claim 1 , wherein the determining key points based on a variation of the radial distances of adjacent points comprises:
setting a threshold for the variation of the radial distances; and
determining at least one of the adjacent points as a key point, wherein the variation of the radial distances of adjacent points is greater than the threshold.
4 . The method of claim 3 , wherein the setting a threshold for the variation comprises setting the threshold based on an angular resolution of the LiDAR.
5 . The method of claim 4 , wherein the setting the threshold based on an angular resolution of the LiDAR comprises:
setting the threshold to be greater than or equal to a product of the angular resolution of the LiDAR and a scanning distance,
wherein the scanning distance is a distance from the laser origin to a detection position.
6 . The method of claim 1 , wherein the filtering out non-ground points from the point cloud based on the key points comprises:
obtaining regions in the top view plane by dividing the top view plane in a radial direction based on the key points;
acquiring height information of the respective regions based on the point cloud; and
determining whether points within the respective regions are non-ground points based on the height information.
7 . The method of claim 2 , wherein the filtering out non-ground points from the point cloud based on the key points comprises:
acquiring, in the top view plane, a plurality of key points on a straight line connecting the coordinate origin with a key point;
obtaining height information of a region between two adjacent key points on the straight line, based on the point cloud; and
determining whether a point in the region is a non-ground point based on the height information.
8 . A device for filtering out non-ground points from a point cloud, comprising:
one or more processors; and
a storage device configured to store one or more programs, that, when executed by the one or more processors, cause the one or more processors to:
acquire a point cloud by scanning the ground with light detection and ranging (LiDAR),
calculate radial distances of adjacent points in the point cloud,
determine key points based on a variation of the radial distances of adjacent points, and
filter out non-ground points from the point cloud based on the key points.
9 . The device of claim 8 , wherein the adjacent points comprise points obtained by scanning the ground using adjacent laser heads of the LiDAR or points obtained by scanning the ground using one laser head of the LiDAR at adjacent scanning angles, wherein the radial distance is a distance from a projection position of a point of the point cloud to a coordinate origin in the top view plane, and the coordinate origin in the top view plane is a projection of a laser origin in the top view plane.
10 . The device of claim 8 , wherein the one or more programs, when executed by the one or more processors, cause the one or more processors further to:
set a threshold for the variation of the radial distances; and
determine at least one of the adjacent points as a key point, wherein the variation of the radial distances of the adjacent points is greater than the threshold.
11 . The device of claim 10 , wherein the one or more programs, when executed by the one or more processors, cause the one or more processors further to set the threshold based on an angular resolution of the LiDAR.
12 . The device of claim 11 , wherein the one or more programs, when executed by the one or more processors, cause the one or more processors further to:
set the threshold to be greater than or equal to a product of the angular resolution of the LiDAR and a scanning distance,
wherein the scanning distance is a distance from the laser origin to a detection position.
13 . The device of claim 8 , wherein the one or more programs, when executed by the one or more processors, cause the one or more processors further to:
obtain regions in the top view plane by dividing the top view plane in a radial direction based on the key points;
acquire height information of the respective regions based on the point cloud; and
determine whether points within the respective regions are non-ground points based on the height information.
14 . The device of claim 9 , wherein the one or more programs, when executed by the one or more processors, cause the one or more processors further to:
acquire, in the top view plane, a plurality of key points on a straight line connecting the coordinate origin with a key point;
obtain height information of a region between two adjacent key points on the straight line based on the point cloud; and
determine whether a point in the region is a non-ground point based on the height information.
15 . A non-transitory computer readable storage medium, in which a computer program is stored, wherein the program, when executed by a processor, causes the processor to execute a method for filtering out non-ground points from a point cloud, comprising:
acquiring a point cloud by scanning ground with light detection and ranging (LiDAR);
calculating radial distances of adjacent points in the point cloud;
determining key points based on a variation of the radial distances of adjacent points; and
filtering out non-ground points from the point cloud based on the key points.