Segmentation processing of image data for LiDAR-based vehicle tracking system and method

1. A method of LiDAR-based vehicle tracking for a non-transitory computer readable storage medium storing one or more programs, the one or more programs comprising instructions, which when executed by a computing device, cause the computing device to perform the following steps comprising:

projecting points in a processed point cloud to a unit sphere;

for each of the projected points, determining that a distance between one and each of its neighboring points is smaller than a threshold and forming a point cluster of the one and its neighboring points, so that each point cluster represents a vehicle on a road;

forming cluster pairs from the point clusters;

calculating a probability that a cluster pair belongs to a same vehicle using tracking information from a previous frame and spatial information from a current frame;

for each cluster pair, determining whether the clusters of the cluster pair are to be merged into a single cluster.

2. The method according to claim 1 , before projecting points, further comprising:

obtaining a raw point cloud from a LiDAR; and

removing points that belong to ground.

3. The method according to claim 2 , wherein removing points further comprises:

determining points that belong to the ground in a LiDAR scan.

4. The method according to claim 1 , after projecting points, further comprising:

forming clusters of the processed points based on Euclidean distance.

5. A system for LiDAR-based vehicle tracking, the system comprising:

an internet server, comprising:

an I/O port, configured to transmit and receive electrical signals to and from a client device;

a memory;

one or more processing units; and

one or more programs stored in the memory and configured for execution by the one or more processing units, the one or more programs including instructions for:

projecting points in a processed point cloud to a unit sphere;

for each of the projected points, determining that a distance between one and each of its neighboring points is smaller than a threshold and forming a point cluster of the one and its neighboring points, so that each point cluster represents an object;

forming cluster pairs from the point clusters;

calculating a probability that a cluster pair belongs to a same object using a time-space representation of a single object;

for each cluster pair, determining whether the clusters of the cluster pair are to be merged into a single cluster.

6. The system according to claim 5 , before projecting points, further comprising:

obtaining a raw point cloud from a LiDAR; and

removing points that belong to ground.

7. The system according to claim 6 , wherein removing points further comprises:

determining points that belong to the ground in a LiDAR scan.

8. The system according to claim 5 , after projecting points, further comprising:

forming clusters of the processed points based on Euclidean distance.

9. The system of claim 5 , wherein the time-space representation of the single object comprises tracking information from a previous frame and spatial information from a current frame.

10. The method of claim 1 , wherein using the spatial information from the current frame comprises constructing a first table using the spatial information from clusters in the current frame, the first table comprising elements p_1(i, j) representing a spatial probability of clusters i and j belonging to the same vehicle.

11. The method of claim 10 , wherein using the tracking information from the previous frame comprises constructing a second table using the tracking information from previous frames, wherein each element p_2(i, j) in the second table represents a tracking probability of clusters I and j belonging to the same vehicle.

12. The method of claim 11 further comprising constructing a third table that is an element-wise multiplication of the first table and the second table, wherein each element p_3(i, j) in the third table represents a probability that clusters i and j belong to the same vehicle.

13. The method of claim 12 , wherein the determining whether the clusters of the cluster pair are to be merged into a single cluster comprises merging the clusters i and j when a probability p_3(i, j) in the third table is greater than a predetermined threshold.

14. The system of claim 5 further comprising at least one LiDAR.

15. The system of claim 5 further comprising a mapping module and a global positioning system (GPS), wherein the one or more programs stored in the memory further comprise instructions for:

using the mapping module and the GPS to obtain information on location of a drivable road; and

determining whether each point cluster is considered to be part of a vehicle.

16. The system of claim 5 further comprising a mapping module and a global positioning system (GPS), wherein the one or more programs stored in the memory further comprise instructions for:

using the mapping module and the GPS to obtain information on location of a drivable road; and

removing any point clusters determined to be not on the drivable road.

17. The system of claim 5 further comprising:

an inertial measurement unit and global positioning system module configured to collect data in an environment; and

a LiDAR configured to provide a raw point cloud to the one or more processing units.

18. The system of claim 15 further comprising a LiDAR configured to provide a raw point cloud to the one or more processing units.