Mobile Robot Position Control Using Computer Vision

Abstract

This paper describes computer vision for controlling board cutter mobile robot position in a known environment. This technique develops camera calibration to control the position of the robot that is directed at a certain position in a known wide area. The system calculate x, y, and orientation of mobile robot position from pixel coordinate to world scale coordinate by changing the pixel scale displayed on the monitor screen in millimeters using camera calibration, therefore the actual position in world coordinate could be known. Fixed camera is placed above environment to make it available to view mobile robot clearly while the mobile robot placed in work area. To find out the actual position of the robot in world scale units, it can be done by using focal length as one of the camera's intrinsic parameters. The experiment results show that the best technique is measuring the distance between the camera and the work area as far as 1500 millimeters. The developed method gives a command to the controller by serial communication to set the track length and also the direction of the robot as input by user. The experiment results show that the accuracy is in average 5% - 10%.