Moving target detection based on Genetic K-means Algorithm
Duan Xiuman, Sun Guoxia, Yang Tao · 2011
Recently, the research on extracting moving objects from video sequences in the current computer vision applications is very popular. A new method based on genetic k-means algorithm from Gaussian mixture model (GMM) to deal with moving target of video image was purposed in this paper. The described method in this paper is learned from clustering idea, which is very different from traditional way. According to Genetic K-means Algorithm, by clustering of pixels on the timeline, the background could be described through several clusters. On the basic of these clusters, moving target detection could be done. Simulation experiment showed that the method could give a good result for the moving target detection.