Kernel based method for surface estimation using Laser scanner data

Accurate surface estimation is critical for autonomous robot navigation on the rough terrain. In this paper, we present a new method for estimating the surface of an unknown arbitrarily shaped terrain from the range data. This problem is generally formulated as the estimation of a function whose zero-set corresponds to the surface. The range data are obtained from indigenous designed Laser range scanner (LRS). The range data obtained from a Laser range scanner sample the terrain surface unevenly, rapidly becoming sparse at longer distances. This paper describes the surface estimation problem as a max-margin based formulation and solved objective function using a kernel-based sub-gradient method. Unlike other methods, we use the additional geometric ray based information to eliminate the unnecessary bump of the surface and increase the precision. Finally, the experimental results validate the robustness of the proposed approach.