摘要
针对随着阈值数量增加,导致传统的多级阈值处理方法已不能满足实时应用的状况,在分析群体智能算法和模式搜索的机理后,提出了一种用于多级图像阈值的基于模式搜索策略的改进IPS算法,以提高群体智能算法在多阈值搜索中的综合性能。该策略是应用步长固定模式搜索算法,在每次迭代中对IPS算法获得的全局历史最优解进行轴向精细搜索,从而从众多次有解中找到最优阈值。仿真实验结果表明,改进IPS的算法不仅具有良好的全局探索和局部优化能力,而且在分别基于BCV方法和KE方法的多级阈值处理方面具有优越的综合性能。
For the multilevelthreshold method of image,traditional threshold methods can no longer meet the requirements of realtime applications with the increasing of threshold-numbers.After analysing the mechanism of swarm intelligence algorithm and pattern search,an improved PS(IPS for short)algorithm based on pattern search strategy for multilevel image threshold was proposed to improve the comprehensive performance of swarm intelligence algorithm in multithreshold search.The strategy applies a step-size fixed-mode search algorithm,and performs an axial fine search on the global historical optimal solution obtained by the IPS algorithm in each iteration,thereby finding an optimal threshold from among many secondary solutions.The simulation results show that the IPS algorithm not only has better global exploration and local optimization,but also has superior comprehensive performance in multilevel threshold processing based on BCV method and KE method respectively.
作者
许韫韬
郭锦
李晓艳
董绵绵
吕志刚
李亮亮
XU Yuntao;GUO Jin;LI Xiaoyan;DONG Mianmian;LYU Zhigang;LI Liangliang(School of Electronics and Information Engineering, Xi’an Technological University, Xi’an 710021, China)
出处
《机械与电子》
2020年第7期7-13,共7页
Machinery & Electronics
基金
西安工业大学大学校长基金项目(XAGDXJJ17012)
陕西省科技厅重点研发计划(2019GY-022)
陕西省教育厅专项科研计划项目(17JK0363)
陕西省组合与智能导航重点实验室项目(SKLIIN-20180201)。
关键词
群体算法
多阈值
模式搜索
布谷鸟算法
图像分割
模式识别
swarm algorithms
multi-threshold
pattern search
Cuckoo algorithm
image segmentation
pattern recognition
