Mobile LiDAR platforms for vehicle tracking

A system comprising:

• a vehicle-mounted light detection and ranging (LiDAR) sensor;
• a vehicle-mounted global positioning system (GPS) receiver;
• a computing node operatively coupled to the LiDAR sensor and GPS receiver, the computing node comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor of the computing node to cause the processor to:
• receive a time-series of point clouds from the LiDAR sensor;
• project each point cloud of the time-series of point clouds onto a two-dimensional plane, the two-dimensional plane being a top-view plane;
• provide the projected point clouds to a convolutional neural network;
• receive from the convolutional neural network at least one bounding box of a vehicle within each of the projected point clouds, wherein receiving the at least one bounding box of the vehicle within each of the projected point clouds comprises:
• determine a first bounding box in a first point cloud of the time-series of point clouds;
• assign a vehicle identifier to the first bounding box, the vehicle identifier associated with the vehicle;
• determine a second bounding box in a second point cloud that intersects the first bounding box in the first point cloud, the second point cloud being immediately after the first point cloud in the time series of point clouds;
• and
• assign the vehicle identifier to the second bounding box;
• determine from the received at least one bounding box a relative speed of the vehicle;
• determine from the GPS receiver a reference speed of the LiDAR sensor;
• and
• from the relative speed and the reference speed, determine an absolute speed of the vehicle.