Aptenodytes Forsteri Optimization Algorithm Based on Adaptive Perturbation of Oscillation and Mutation Operation for Image Multi-Threshold Segmentation

Panli Zhang,Jingnan Yang,Fanfan Lou,Jiquan Wang,Xiaobo Sun

DOI: https://doi.org/10.1016/j.eswa.2023.120058

IF: 8.5

2023-01-01

Expert Systems with Applications

Abstract:Multi-threshold segmentation is widely used in image segmentation, but its computational complexity increases with the number of thresholds, which can compromise segmentation precision. Additionally, existing heuristic methods for multi-threshold segmentation still suffer from slow convergence and premature convergence. To address these limitations, this study proposes a novel Aptenodytes Forsteri optimization algorithm with an adaptive perturbation of oscillation and mutation operation (AFOA-APM) to determine the optimal image threshold. AFOA-APM improves upon existing methods with several key innovations: firstly, an adaptive perturbation of the oscillation strategy (AP) improves exploitation ability; secondly, improved methods for move strategies 2 and 3 reduce time complexity and the probability of stagnation; thirdly, a random mutation strategy keeps diversity and preserves inheritance; and finally, an adaptive selection strategy balances search capabilities. AFOA-APM is validated using CEC 2017 benchmark functions and image segmentation problems, and results demonstrate its superior performance in terms of quality, consistency, and accuracy. These findings indicate that AFOA-APM has a high potential to address image multi-threshold segmentation problems effectively.