Implementation of low cost vision based measurement system: Motion analysis of indoor robot

The current paper focuseson the development of a low cost vision based measurement system for obtaining the position, velocity,and trajectory of a mobile robot in an indoor environment using monocular camera. The developed methodutilizedcolor based feature extraction and blob analysis technique to detect and track the motion of the mobile robot. The equations for transforming pixel coordinate into real world coordinate were generated using polynomial surface fit. The noises in the measured data due to unaccounted source of errors were reduced by the use of median filter. The experiment utilizeda commercial webcamto capture the robot motiontherebyreducing the cost of the system. The experiments conducted on a two dimensional platform of size 1600 mm ? 800 mm verified that the methodcan be effectively implemented for detecting mobile robots.