Tsallis Entropy Threshold Image Segmentation Algorithm Based on Improved Honey Bee Mating Algorithm

Aiming at the shortcomings of the traditional honey bee mating algorithm, an improved honey bee mating algorithm was proposed. Firstly, the custom breeding probability and mutation probability in the traditional honey bee mating algorithm were improved to be adaptive to improve the fast convergence and global optimization ability of the algorithm; Secondly, the elite retention strategy was introduced to shorten the optimization time; Finally, with the non-extensibility of Tsallis entropy, using the Tsallis entropy as the fitness function of threshold image segmentation, and the improved honey bee mating algorithm was used to achieve multi-threshold image segmentation to avoid the non-add-on information of the image segmentation process being ignored, thereby improving the accuracy of image segmentation. In order to verify the feasibility of the proposed algorithm, the algorithm was simulated, and compared with some algorithms. The experimental results showed that the algorithm had further improved the optimization ability, fast convergence ability, image segmentation accuracy, image segmentation speed, and so on.