Applications of the Sparse Hough Transform for laser Data Line Fitting and Segmentation

This paper proposes a novel algorithm called the sparse Hough transform, which is shown to have better performances than the standard Hough transform for sparse input data collected from a laser on a mobile robot. In the context of laser sensing and perception for autonomous ground robots, this paper studies performances of the sparse Hough transform and compares it with other segmentation and fitting algorithms. Pseudo-code for the algorithm, theoretical analysis, computer simulations, hardware experiments, and experimental analysis of the sparse Hough transform are presented.