A hybrid neural control scheme for visual-motor coordination

Visual-motor coordination, also referred to as hand-eye coordination, in the context of robotics is the process of using visual information to control a robot manipulator to reach a target point in its workspace. The task requires learning the mapping that exists between camera output and desired end effector location. Biological organisms have demonstrated their superior adaptive capabilities in motion control over present-day robotic systems. Inspired by this fact, various neural network models based on biological systems have been developed for robot control tasks. The drawback of many neural schemes to tackle visual-motor control problems is that of a long training period. We suggest an approach using Kohonen's self-organizing scheme to learn this hand-eye coordination problem in reduced time with high accuracy. Our approach has also been compared with a conventional calibration-based algorithm. These schemes have been implemented in real time on a CRS PLUS robot arm. Experimental results show that the proposed neural scheme is on an average 10 times faster in training compared to similar neural approaches existing in the literature. This speed is also comparable to the conventional algorithm and is more accurate.

[1]  Roger Y. Tsai,et al.  Techniques for calibration of the scale factor and image center for high accuracy 3D machine vision metrology , 1987, Proceedings. 1987 IEEE International Conference on Robotics and Automation.

[2]  Yuh-Lin Chang,et al.  Adaptive self-calibration of vision-based robot systems , 1989, IEEE Trans. Syst. Man Cybern..

[3]  Michael Kuperstein,et al.  Adaptive visual-motor coordination in multijoint robots using parallel architecture , 1987, Proceedings. 1987 IEEE International Conference on Robotics and Automation.

[4]  Helge J. Ritter,et al.  Three-dimensional neural net for learning visuomotor coordination of a robot arm , 1990, IEEE Trans. Neural Networks.

[5]  Roger Y. Tsai,et al.  A versatile camera calibration technique for high-accuracy 3D machine vision metrology using off-the-shelf TV cameras and lenses , 1987, IEEE J. Robotics Autom..

[6]  T. Kohonen.  Self-organized formation of topographically correct feature maps , 1982 .

[7]  W. Thomas Miller,et al.  Real-time application of neural networks for sensor-based control of robots with vision , 1989, IEEE Trans. Syst. Man Cybern..

[8]  Klaus Schulten,et al.  Topology-conserving maps for learning visuo-motor-coordination , 1989, Neural Networks.

[9]  M Kuperstein,et al.  Neural model of adaptive hand-eye coordination for single postures. , 1988, Science.

[10]  Klaus Schulten,et al.  Implementation of self-organizing neural networks for visuo-motor control of an industrial robot , 1993, IEEE Trans. Neural Networks.