Self-adaptive parameter selection in one-dimensional tsallis entropy thresholding with Particle Swarm Optimization algorithm

A self-adaptive parameter selection algorithm for parameter q in one-dimensional Tsallis entropy image thresholding is presented based on optimization algorithm. The method can get the suitable parameter and the optimal threshold value for different kinds of images, which selects the parameter based on the uniformity measure, an image segmentation quality evaluation criterion, as fitness function and searches for the optimal parameter by the Particle Swarm Optimization (PSO) algorithm. The results show that in general cases the optimal parameter q can be found between 0 and 1; if more accurate segmentation is needed, the optimal value of q can be found between 0 and 10.