Vision sensor based tracking and focusing of 3-D singular points

In this paper, a laser scanning vision sensor based automatic focusing and positioning servo control mechanism is presented for 3-D singular points. The mechanism consists of two parts: position detecting and tracking/focusing control. In particular, difference-from-moving average (DMA) algorithm is used to process the 3-D data from a circular laser scanning vision sensor for detecting the position of singular points. The processed information is then fed back to the servo motion control and actuating system for accurate tracking and focusing of singular points. The result shows that this mechanism achieves desired tracking speed and accuracy with experimental validation on a 3-D vision tracking platform.

[1]  Ramesh Jain,et al.  Techniques For Real-Time, 3D, Feature Extraction Using Range Information , 1985, Other Conferences.

[2]  Kevin G. Harding,et al.  3D laser measurements on scattering and translucent surfaces , 1996, Other Conferences.

[3]  Ernest L. Hall,et al.  Three Dimensional Perception For Robot Vision , 1983, Optics & Photonics.

[4]  Francois Blais,et al.  A Very Compact Real Time 3-D Range Sensor For Mobile Robot Applications , 1989, Other Conferences.

[5]  Murali Subbarao,et al.  Three dimensional machine vision using image defocus , 1994 .

[6]  Zhang Ren,et al.  Algorithm design for a position tracking sensor based on pattern recognition , 2002, IEEE 2002 28th Annual Conference of the Industrial Electronics Society. IECON 02.

[7]  Kevin Harding.  THE PROMISE OF MACHINE VISION , 1996 .

[8]  Xiang Chen,et al.  Fuzzy Associative Database for multiple planar object recognition , 2003, Proceedings of the 2003 International Symposium on Circuits and Systems, 2003. ISCAS '03..