Comparision of Edge Detection Methods for Obstacles Detection in a Mobile Robot Environment

Abstract

This paper presents application of different edge detection methods for detecting of obstacles in the environment of mobile robot. Autonomous mobile robots operating in an unknown environment must be able to cope with hazards and obstacles, and at the same time to navigate through the environment performing required tasks. Detection and avoidance of obstacles, collisions and hazardous situations is a very important part of reliable and secure navigation of mobile robots. Vision sensor (camera) provides large amount of data about obstacles in the environment, however it is necessary to extract only the most important details. This paper will present a comparison of the four most common edge detection methods on examples of obstacles detection in different conditions. Methods used in this study are: Sobel, Prewitt, Roberts and Canny.