A State Transition Algorithm Based on Jump Operator Applied to Image Threshold Segmentation

This paper proposes an improved state transition algorithm (JUSTA) based on the OTSU multi-threshold segmentation algorithm. We introduce a state transition algorithm (STA) into image threshold segmentation and propose a new jump operator to improve the performance of STA according to the characteristics of threshold solution space and a multi-threshold segmentation algorithm. The jump operator enhances the local search ability of the algorithm by adding synchronization characteristics. At the same time, the variation factor modified by quadratic function and the dynamic mutation probability is introduced to improve the ability to jump out of the current state. A large number of experimental studies show that the JUSTA algorithm has obvious advantages in stability and rapidity in finding the best threshold of image threshold segmentation model compared with genetic algorithm (GA), particle swarm optimization (PSO), artificial bee colony algorithm (ABC) and state transition algorithm (STA).