A visual feedback control system for tracking and zooming a target

The authors describe an adaptive sensing system with tracking and zooming for a moving object in a time varying environment. Two algorithms, the close contour mating technique and the effective determination of zoom ratio by fuzzy control, are proposed for achieving the sensing system. The estimation of objective feature parameters, two-dimensional velocity and size, is based on close contour matching. First, the correspond problem is solved by using cross-correlation in projection which is abstracted from the contours of the object in the specialized difference image. Closely shifting and matching the contours eliminates the occluded contours as well as noise. To zoom the tracked object in accordance with the state of its shape or movement, fuzzy control is considered. Three sets of input membership functions, the confidence of object shape, the variance of object velocity, and the object size, are evaluated based on fuzzy rules for controlling the focus length of the sensor. Experimental results show that the proposed system is robust and valid for numerous kinds of moving objects in real scenes with a system period of 1.85 s