Research on multi-objective automatic tracking algorithm for vision-guide robot

To ensure the multi-objective automatic tracking robustness, a new algorithm is proposed. The motion image restoration is got by adaptive Wiener filtering algorithms based on the image fuzzy analysis for motion object. The object tracking motion state model is designed with motion object parameter. The tracking template updating real-time is realized by Kalman motion parameter prediction algorithm with error estimation, so it can track multi-objective automatically. Experimental results demonstrated that the proposed tracking algorithm is not affected by illumination and noisy, and it meets the practical application requirement. Copyright © 2011 Binary Information Press.