Terrain classification for mobile robot navigation

Abstract

Terrain classification is an important problem that still remains to be solved for autonomous robot vehicle guidance. Often, obstacle detection systems are used which cannot distinguish between solid obstacles such as rocks or soft obstacles such as tall patches of grass. It can also be used to recognize sand roads or other drivable areas. The aim of this project is to implement a terrain classification algorithm. The algorithm comprises of two main modules - color thresholding module and a neural network module. The initial input terrain image with distinct features namely sky, road, tree blob is given as an input to the color thresholding module. The input stream of images is first preprocessed by resizing the images for standardizing the images to a standard image space.