4D scan registration with the SR-3000 LIDAR

An algorithm and data management scheme is presented to utilize the data output of the SR-3000, a 3D LIDAR sensor. The SR-3000 generates a 4D point cloud at video frame rates where each point is described by its 3D coordinates and its optical intensity. The implementation of each major component of the algorithm is described, including a projective ICP algorithm, reverse calibration equations for the SR-3000 sensor, the processing of the intensity data, and a point selection algorithm which includes the use of a 3D volume feature extraction algorithm. This ICP based alignment algorithm is used in the context of a data management scheme that stores and retrieves scan data and intermediate data products to realize pseudo-global scan alignment, data compression, and real time data display.