Abstract：Robot dynamic hand-eye coordination based on online recognition is presented in this paper.Through dynamic motion planning using the results of online recognition of objects,the method improves the flexibility and the adaptability of the hand-eye system.Experimental results show that the recognition algorithms in this paper have perfect real-time performance and reliability.The man-machine interface is improved by introducing voice control.
[1] Hervé J,Sharma R,Cucka P.Toward Robust Vision-based Control:Hand-Eye Coordination without Calibration.Proceedings of 1991 IEEE International Symposium on Intelligent Control,1991:457-462 [2] Yoshimi B H,Allen P K.Active Uncalibrated Visual Servoing.Proceedings of 1994 IEEE International Conference on Robotics and Automation,1994:156-161 [3] Hager G D,Chang W C,Morse A S.Robot Feedback Control Based on Stereo Vision:Towards Calibration-free Hand/eye Coordination.Proceedings of 1994 IEEE International Conference on Robotics and Automation,1994:2850-2856 [4] Tsai R Y.An Efficient and Accurate Camera Calibration Technique for 3D Machine Vision.Proceedings of CVPR'86:364～374 [5] Kenneth R Castleman.Digital Image Processing.Prentice Hall,Upper Saddle River,New Jersey,1996 [6] 余松煜等.数字图像处理.北京:电子工业出版社.1987 [7] 钟玉琢等.机器人视觉技术.北京:国防工业出版社.1994
