Online robot tracking using genetic algorithms

摘要

This paper presents the applicability of using a genetic algorithm (GA) for robot tracking inside a predefined environment in real time. The GA is used to determine the position of a moving robot inside a soccer robot field. A proposed new crossover algorithm is presented. A video file and an image containing the robot's physical appearance (robot template) are inputs to the system. Experimental results show that GA is able to locate the robot.