Robust shape-matching criterion for precision tracking of moving objects in correlated image sequences

Abstract

A robust shape-matching measure with a boundary detection preprocessor is proposed to track a moving object in complex image sequences. A correlation-tracking algorithm seeks to align the incoming target image with the reference image of the target. But it has a critical problem if we use the conventional block-matching criteria. This is the so-called 'false peak problem', which is generally caused by highly correlated background pixels with similar intensities to the pixels of moving targets. The robustness of the proposed method for practical application is demonstrated by simulating two kinds of real-target image sequences.