| 摘 要: 针对非洲秃鹫算法(AVOA)全局搜索能力不足与局部搜索策略冗杂的缺点,提出一种改进非洲秃鹫算法(β-PAVOA)。算法采用分段线性混沌映射(PWLCM)初始化种群,增强种群多样性。引入β 分布与基于饥饿率的搜索策略,增强算法全局搜索能力。改进原算法局部搜索策略,帮助算法及时跳出局部最优。通过8个测试函数验证算法的有效性,并将其应用于二维Otsu图像阈值分割模型。实验结果表明,在测试函数上,β-PAVOA相比较于非洲秃鹫算法(AVOA)、金豺狼优化算法(GJO)、灰狼算法(GWO)、鲸鱼优化算法(WOA)和粒子群算法(PSO)有着更好的精度与收敛速度;在二维Otsu图像阈值分割模型上,β-PAVOA在搜到最优解的情况下收敛速度也仍然领先,这也证明了改进算法的有效性。 |
| 关键词: 元启发式算法;改进非洲秃鹫优化算法;β 分布;分段线性混沌映射;图像阈值分割 |
|
中图分类号: TP311
文献标识码: A
|
| 基金项目: 江苏省研究生实践创新计划项目(SJCX22-XZ033,SJCX23-XY069,SJCX23-XY071);2023年大学生创新创业训练计划项目(2023591,2023576) |
|
| Research on Image Threshold Segmentation Based on Improved African Vulture Optimization Algorithm |
|
ZHANG Xiaoyu, FANG Zhongqing, DU Yi, KONG Weibin, WANG Yuting, CHENG Ziyao
|
(School of Information Engineering, Yancheng Institute of Technology, Yancheng 224051, China)
xiaoyuzhang721@163.com; fangzq@ycit.edu.cn; Lemondu213@163.com; kongweibin@ycit.cn; 2251193251@qq.com; 3417747952@qq.com
|
| Abstract: Aiming at the shortcomings of the African Vulture Optimization Algorithm (AVOA) in terms of insufficient global search capability and redundant local search strategies, this paper proposes an improved African Vulture Optimization Algorithm (β-PAVOA). Piecewise Linear Chaotic Map (PWLCM) is used to initialize population and enhance population diversity. The β-distribution and hunger-rate-based search strategy are introduced to enhance the global search capability of the algorithm. The local search strategy of the original algorithm is improved to help the algorithm jump out of the local optimum in time. The effectiveness of the algorithm is verified by eight test functions and it is applied to the two-dimensional Otsu image threshold segmentation model. The experimental results indicate that β-PAVOA has better accuracy and convergence speed on the test functions, compared to the African Vulture Algorithm ( AVOA), Golden Jackal Optimization ( GJO), Grey Wolf Optimization ( GWO), Whale Optimization Algorithm ( WOA), and Particle Swarm Optimization ( PSO). On the two-dimensional Otsu image threshold segmentation model, β-PAVOA still leads in convergence speed even when the optimal solution is found, which also proves the effectiveness of the improved algorithm. |
| Keywords: metaheuristic algorithm; improved African Vulture Optimization Algorithm; β distribution; PWLCM; image threshold segmentation |
