An Adaptive Bi-Mutation-Based Differential Evolution Algorithm for Multi-Threshold Image Segmentation

Aiming at solving the problems of large calculation, time-consuming, and low segmentation accuracy multi-threshold image segmentation, an adaptive threshold value based differential evolution algorithm is proposed in this paper. Firstly, opposite learning strategy introduced into initial population to improve quality population; secondly, a threshold-value-based mutation balance exploration development capabilities algorithm, number successfully evolved individuals considered as adaptively adjust superior inferior individuals. Experiments demonstrate that has better performance enhancing speeding up convergence.