System and method for providing online multi-LiDAR dynamic occupancy mapping
A system and method for providing online multi-LiDAR dynamic occupancy mapping that include receiving LiDAR data from each of a plurality of LiDAR sensors. The system and method also include processing a region of interest grid to compute a static occupancy map of a surrounding environment of the ego vehicle and processing a dynamic occupancy map. The system and method further include controlling the ego vehicle to be operated based on the dynamic occupancy map.
1. A computer-implemented method for providing online multi-LiDAR dynamic occupancy mapping, comprising:
receiving LiDAR data from each of a plurality of LiDAR sensors, wherein the LiDAR data is associated with a surrounding environment of an ego vehicle;
processing a region of interest grid to compute a static occupancy map of the surrounding environment of the ego vehicle;
processing a dynamic occupancy map, wherein phase congruency is utilized to segment dynamic and static objects that are determined based on the static occupancy map, wherein a phase congruency score is computed from motion vectors that are determined from cells of the dynamic occupancy map; and
controlling the ego vehicle to be operated based on the dynamic occupancy map.
2. The computer-implemented method of claim 1 , wherein processing the region of interest grid includes receiving the region of interest grid from a region of interest grid pool.
3. The computer-implemented method of claim 2 , wherein processing the region of interest grid includes analyzing the LiDAR data received from each of the plurality of LiDAR sensors and processing the region of interest grid for each of the plurality of LiDAR sensors.
4. The computer-implemented method of claim 3 , wherein the processed region of interest grid includes cells which partition at least one space of the surrounding environment of the ego vehicle into finitely fixed sized grids.
5. The computer-implemented method of claim 3 , wherein processing the region of interest grid includes storing the processed region of interest grid within a region of interest grid buffer to be utilized in global occupancy and grid update processing.
6. The computer-implemented method of claim 5 , wherein point coordinates associated with the LiDAR data are transformed from a sensor frame to an ego vehicle frame, wherein transformed point coordinates are distributed into region of interest cells based on respective two-dimensional LiDAR coordinates.
7. The computer-implemented method of claim 6 , further including analyzing a height difference between a highest and lowest point in each region of interest cell against a predetermined obstacle threshold to determine if the height difference is greater than the predetermined obstacle threshold, wherein the region of interest cells is marked as an obstacle cell when the height difference is greater than the predetermined obstacle threshold and the static occupancy map is computed.
8. The computer-implemented method of claim 7 , wherein processing the dynamic occupancy map includes updating motion cells based on updating obstacle probabilities of region of interest cells, wherein each motion cell includes a grid of occupancy cells.
9. The computer-implemented method of claim 8 , wherein the phase congruency score is determined for each motion cell of the dynamic occupancy map from a combination of the motion vectors with previously obtained motion vectors obtained at pervious iterations, wherein the phase congruency score is compared against a predetermined phase congruency threshold to denote at least one cell as a dynamic cell.
10. A system for providing online multi-LiDAR dynamic occupancy mapping, comprising:
a memory storing instructions when executed by a processor cause the processor to:
receive LiDAR data from each of a plurality of LiDAR sensors, wherein the LiDAR data is associated with a surrounding environment of an ego vehicle;
process a region of interest grid to compute a static occupancy map of the surrounding environment of the ego vehicle;
process a dynamic occupancy map, wherein phase congruency is utilized to segment dynamic and static objects that are determined based on the static occupancy map, wherein a phase congruency score is computed from motion vectors that are determined from cells of the dynamic occupancy map; and
control the ego vehicle to be operated based on the dynamic occupancy map.
11. The system of claim 10 , wherein processing the region of interest grid includes receiving the region of interest grid from a region of interest grid pool.
12. The system of claim 11 , wherein processing the region of interest grid includes analyzing the LiDAR data received from each of the plurality of LiDAR sensors and processing the region of interest grid for each of the plurality of LiDAR sensors.
13. The system of claim 12 , wherein the processed region of interest grid includes cells which partition at least one space of the surrounding environment of the ego vehicle into finitely fixed sized grids.
14. The system of claim 12 , wherein processing the region of interest grid includes storing the processed region of interest grid within a region of interest grid buffer to be utilized in global occupancy and grid update processing.
15. The system of claim 14 , wherein point coordinates associated with the LiDAR data are transformed from a sensor frame to an ego vehicle frame, wherein transformed point coordinates are distributed into region of interest cells based on respective two-dimensional LiDAR coordinates.
16. The system of claim 15 , further including analyzing a height difference between a highest and lowest point in each region of interest cell against a predetermined obstacle threshold to determine if the height difference is greater than the predetermined obstacle threshold, wherein the region of interest cells is marked as an obstacle cell when the height difference is greater than the predetermined obstacle threshold and the static occupancy map is computed.
17. The system of claim 16 , wherein processing the dynamic occupancy map includes updating motion cells based on updating obstacle probabilities of region of interest cells, wherein each motion cell includes a grid of occupancy cells.
18. The system of claim 17 , wherein the phase congruency score is determined for each motion cell of the dynamic occupancy map from a combination of the motion vectors with previously obtained motion vectors obtained at pervious iterations, wherein the phase congruency score is compared against a predetermined phase congruency threshold to denote at least one cell as a dynamic cell.
19. A non-transitory computer readable storage medium storing instructions that when executed by a computer, which includes a processor perform a method, the method comprising:
receiving LiDAR data from each of a plurality of LiDAR sensors, wherein the LiDAR data is associated with a surrounding environment of an ego vehicle;
processing a region of interest grid to compute a static occupancy map of the surrounding environment of the ego vehicle;
processing a dynamic occupancy map, wherein phase congruency is utilized to segment dynamic and static objects that are determined based on the static occupancy map, wherein a phase congruency score is computed from motion vectors that are determined from cells of the dynamic occupancy map; and
controlling the ego vehicle to be operated based on the dynamic occupancy map.
20. The non-transitory computer readable storage medium of claim 19 , wherein the phase congruency score is determined for each motion cell of the dynamic occupancy map from a combination of the motion vectors with previously obtained motion vectors obtained at pervious iterations, wherein the phase congruency score is compared against a predetermined phase congruency threshold to denote at least one cell as a dynamic cell.