GEOMETRIC DRIVABLE SURFACE ESTIMATION
A point cloud generated at least in part using a lidar sensor is received. A geometric mesh based on the point cloud is determined. A seed geometric face formed in the geometric mesh is selected based on one or more seed selection criteria. Starting from the seed geometric face, neighboring geometric faces of the geometric mesh that meet one or more relative neighbor selection criteria are iteratively selected into a region group, and an operable region indicated by the region group is detected.
1 . A method, comprising:
receiving a point cloud generated at least in part using a lidar sensor;
determining a geometric mesh based on the point cloud;
selecting a seed geometric face formed in the geometric mesh based on one or more seed selection criteria;
starting from the seed geometric face, iteratively selecting neighboring geometric faces of the geometric mesh that meet one or more relative neighbor selection criteria into a region group; and
detecting an operable region indicated by the region group.
2 . The method of claim 1 , wherein the geometric mesh is generated based on a two-dimensional flat geometric mesh.
3 . The method of claim 2 , wherein the two-dimensional flat geometric mesh is determined based on azimuth and altitude values of the point cloud.
4 . The method of claim 2 , wherein the two-dimensional flat geometric mesh is determined independent from distance values associated with the point cloud.
5 . The method of claim 1 , wherein the geometric mesh is a three-dimensional mesh, and wherein the geometric mesh includes azimuth, altitude, and distance dimensions.
6 . The method of claim 1 , wherein the geometric mesh includes a plurality of triangles, wherein each vertex of the plurality of triangles corresponds to a point of the point cloud.
7 . The method of claim 1 , wherein the one or more seed selection criteria includes determining a normal of a geometric face of the geometric mesh and comparing the normal to a reference axis of the lidar sensor.
8 . The method of claim 1 , wherein selecting the seed geometric face formed in the geometric mesh based on the one or more seed selection criteria includes determining a height for a candidate geometric face.
9 . The method of claim 8 , wherein the height for the candidate geometric face is based on a corresponding altitude value for each vertex of the candidate geometric face.
10 . The method of claim 1 , wherein the one or more relative neighbor selection criteria includes determining a normal of a geometric face of the geometric mesh.
11 . The method of claim 10 , further comprising comparing the normal of the geometric face of the geometric mesh to a reference axis.
12 . The method of claim 10 , further comprising comparing the normal of the geometric face of the geometric mesh to a normal of a reference geometric face.
13 . The method of claim 12 , wherein the reference geometric face is included in the seed geometric face.
14 . The method of claim 1 , wherein the detected operable region is associated with a drivable surface.
15 . The method of claim 1 , further comprising associating a drivable characteristic with at least one vertex of the geometric mesh.
16 . The method of claim 1 , further comprising associating a drivable characteristic with at least one geometric face of the geometric mesh.
17 . The method of claim 1 , further comprising removing a face from the geometric mesh based on a property of the face.
18 . The method of claim 17 , wherein the property of the face is a length of an edge of the face, an area of the face, an intensity value associated with the face, or a value of an angle of the face.
19 . A system, comprising:
a lidar sensor to provide one or more measurements of targets located downrange from the system;
one or more processors; and
a memory coupled to the one or more processors, wherein the memory is configured to provide the one or more processors with instructions which when executed cause the one or more processors to:
receive a point cloud generated at least in part using the lidar sensor;
determine a geometric mesh based on the point cloud;
select a seed geometric face formed in the geometric mesh based on one or more seed selection criteria;
starting from the seed geometric face, iteratively select neighboring geometric faces of the geometric mesh that meet one or more relative neighbor selection criteria into a region group; and
detect an operable region indicated by the region group.
20 . A computer program product, the computer program product being embodied in a non-transitory computer readable storage medium and comprising computer instructions for:
receiving a point cloud generated at least in part using a lidar sensor;
determining a geometric mesh based on the point cloud;
selecting a seed geometric face formed in the geometric mesh based on one or more seed selection criteria;
starting from the seed geometric face, iteratively selecting neighboring geometric faces of the geometric mesh that meet one or more relative neighbor selection criteria into a region group; and
detecting an operable region indicated by the region group.