A fast self-adapt target image segmentation algorithm

Abstract

On the base of 2D maximum between-cluster variance algorithm, a fast self-adapt target image segmentation algorithm is brought forward. The algorithm utilizes not only the gray level information of each pixel and its spatial correlation information within the neighborhood, but also the dimension of neighborhood domain. Because the traditional 2D maximum between-cluster variance algorithm spends a lot of time on circulating operation and because genetic algorithm can search best value in the whole field, the mix-genetic algorithm is put into the image segmentation algorithm in which the gray level, neighborhood dimension and average value are encoded. This method can quicken the speed of producing the best threshold value, improve greatly the performance of the methods. The experiment results show that the self-adapt image segmentation algorithm spends more short time on running and has better segmentation effect than the traditional algorithm, so the self-adapt segmentation algorithm has the image real-time segmentation application merit.