Optimal path planning for planetary exploration rovers based on artificial vision system for environment reconstruction

In this paper an optimal path planning algorithm for rover navigation is presented. The algorithm is based on an artificial vision system which reconstructs the digital elevation map of the environment through the use of two stereo cameras. On the basis of this digital elevation map the optimal path is generated with a simulated annealing approach. The direct use of the information in the digital elevation map allows to design a detailed path which takes care of each detected feature of the environment and of the cinematic limitation of the rover