A Motion Estimation Algorithm Based on the Wavelet Domain
一种基于小波域的运动估计方法

This paper describes a novel motion object detection and tracking algorithm based on the Multiscale Edges Detection (MED) and Continuous Wavelet Transform (CWT). Multiscale representations are very effective for analyzing the information content of images. First, moving object multiscale edges are detected and characterized from the local maxima of a wavelet transform. Then, the CWT is used to map the input signal space to a motion parameters space. It is designed so as to possess simultaneously the properties of MED and CWT. Experimental results show that the proposed algorithm is effective with robustness for noise, rotation and occlusions.