A 2/sup 1///sub 2/-D algorithm for the shortest path search running on 2-D hardware

Abstract

Presents an algorithm for individuating the shortest path in a 3D space in which the obstacles are described by means of range data. Such an algorithm runs on 2D hardware which is normally used for image processing tasks avoiding the development of expensive 3D hardware. This is possible because the procedure is split in two different 2D processes. Main advantages of the technique are: no pre-processing of the data coming from the sensors is needed, the algorithm works directly on them; and the computation complexity is linear with the length of the path.<>