Iterative Closest Point registration algorithm based on intensity feature matching of TOF point cloud

Abstract:

There are many noise points in the 3D point cloud dala obtained by Time Of Flight (TOF) and the proportion of real objects in the point cloud is small. An iterative closest point registration algorithm based on intensity feature matching is proposed for TOF point cloud data. Firstly, the intensity feature is employed to extract the effective region, and then the effective region is configured. Finally, the change matrix of the effective region is utilized to register the whole point cloud data. Experimental results show that this method can effectively improve the registration accuracy of real target point cloud without affecting the registration speed.