Navigation of mobile robots using artificial intelligence technique.

Abstract

The ability to acquire a representation of the spatial environment and the ability to localize within it are essential for successful navigation in a-priori unknown environments. This document presents a computer vision method and related algorithms for the navigation of a robot in a static environment. Our environment is a simple white colored area with black obstacles and robot (with some identification mark-a circle and a rectangle of orange color which helps in giving it a direction) present over it. This environment is grabbed in a camera which sends image to the desktop using data cable. The image is then converted to the binary format from jpeg format using software which is then processed in the computer using MATLAB. The data acquired from the program is then used as an input for another program which controls the robot drive motors using wireless controls. Robot then tries to reach its destination avoiding obstacles in its path. The algorithm presented in this paper uses the distance transform methodology to generate paths for the robot to execute. This paper describes an algorithm for approximately finding the fastest route for a vehicle to travel one point to a destination point in a digital plain map, avoiding obstacles along the way. In our experimental setup the camera used is a SONY HANDYCAM. This camera grabs the image and specifies the location of the robot (starting point) in the plain and its destination point. The destination point used in our experimental setup is a table tennis ball, but it can be any other entity like a single person, a combat unit or a vehicle.