Dynamic image analysis (DIA) is a very active branch in the image processing and computer vision field. It spans many technologies including computer science, machine vision, image engineering, pattern analysis, artificial intelligence, etc. and attracts large numbers of researchers' attention. The basic task of DIA is detecting the motion information from image sequences, constructing motion model, recognizing and tracking the moving targets.As to Dynamic image analysis, there are still many problems both in theory research and in applications, due to the complexity of detecting, modeling and tracking the moving targets and the restriction of video technology. Large numbers of researchers have been devoting themselves in the area and have already achieved many progresses. The dissertation studies the key technologies of DIA based on these achievements. The main contributions of this dissertation can be summarized as follows:Firstly, image pretreatment technologies are investigated. A wavelet-based algorithm for image contrast enhancement is proposed. The approach treats the correlation between wavelet planes as providing an indication of the likelihood that noise is present, then increase wavelet coefficients due to signal and suppress wavelet coefficients due to noise. The algorithm solves the problem of noise amplification while enhance signal visibility.Secondly, moving target detection in dynamic scenes is introduced. The method extracts the contour of each moving object based on the fusion of a motion segmentation technique using image subtraction and region growing process, and then the final result after the optimization with active contour model can be achieved. The main works in the detection are as follows: a converse watershed algorithm is proposed, which gets the number of the moving targets and their position with the clustering of the image subtraction; A coincidence indicator is...
|