Construction of 3D Map in Robot by Combining Action and Perceived Images

In this paper, we propose an active vision strategy for the construction of 3D map in a robot brain from its stereo eye images. We show that a combination of the robot action and image change caused by the action will improve the accuracy of the vision system parameters. If the robot stereo cameras have been accurately calibrated, the obtained reconstruction of the 3D static scene stays stationary. If not accurate, the reconstructed scene changes according to the robot action. Therefore, we utilize this image change to correct the parameters of the stereo cameras so as to obtain stationary scene under the action. We show the feasibility of this idea as an auto-calibration of robot vision with some experimental simulations and implementation on an actual robot.