Building three dimensional maps based on cellular neural networks

Abstract

One of the main problem in robotics is map building. In the article 3D map building based on 3D laser data is presented. The data is obtained from a SICK laser mounted on a rotating support what gives a 3D representation of the scene. The map is divided into cells and each cell represents a certain area of the scene and keeps a list of objects. This is a real-time system, which consumes little computer memory and works properly in indoor environment.