Maximum entropy multi-threshold image segmentation based on improved particle swarm optimization

Abstract

Abstract To solve the problem of low speed and low precision of traditional maximum entropy image segmentation algorithm, a multi-threshold image segmentation algorithm based on improved particle swarm optimization algorithm is proposed. Taking the optimal threshold optimization problem in multi-threshold image segmentation as the research object, the optimal objective function is obtained by using the maximum entropy multi-threshold segmentation method, and then the maximum entropy method and particle swarm optimization algorithm are fused. In order to solve the problem that particle swarm optimization (PSO) is prone to fall into local optimization in the later iteration process, the PSO is improved and the expansion model is added. Finally, the maximum entropy multi-threshold image segmentation method based on the standard particle swarm optimization algorithm and the maximum entropy multi-threshold image segmentation method based on the improved particle swarm optimization algorithm segment the h-component image (hue, saturation, value) of HSV. Images are converted from images. The segmentation results of the two algorithms are evaluated by running time, and the structural similarity of the algorithms is evaluated. evaluation result. Experimental results show that the improved maximum entropy multi-threshold image segmentation algorithm based on particle swarm optimization can better achieve complex image segmentation, and the algorithm has stronger real-time performance.