摘要
基于三维直方图的最大类间方差阈值法(三维Otsu)考虑了邻域均值和中值信息,抗噪性能较好,可以获得理想的分割结果,然而其计算复杂度非常高,效率低下。萤火虫算法(Firefly Algorithm,FA)是一种新型的启发式算法。本文在介绍萤火虫算法基本原理的基础上,提出一种基于莱维飞行的分簇萤火虫算法(CBLFA),并用于改进三维Otsu阈值法的效率。实验结果表明该方法可以快速获得适合的阈值,适应度函数值总体上优于基本萤火虫算法和基本粒子群算法,是一种鲁棒性更强的三维Otsu阈值分割法。
Otsu based on the 3D histogram （3D Otsu） considers the neighborhood mean and median in- formation, it can obtain the ideal segmentation results and has better anti -noise performance; however, its computational complexity is high and computational efficiency is very low. Firefly algorithm （FA） is a newly proposed meta - heuristics algorithm. After the fundamental of FA is illustrated, the clustering fire- fly algorithm based on levy flight （CBLFA） is proposed and utilized for improve 3D Otsu. Experimentalresuhs show that CBLFA can quickly obtain the suitable thresholds, significantly reduce the execution time and could obtain the better fitness function value than basic FA and particle swarm algorithm on the whole.
作者
叶志伟
徐炜
赵伟
侯玉倩
杨娟
YE Zhiwei XU Wei ZHAO Wei HOU Yuqian YANG Juan(Shool of Computer Science, Hubei University of Technology, Wuhan 430068, China)
出处
《中国体视学与图像分析》
2016年第4期374-380,共7页
Chinese Journal of Stereology and Image Analysis
基金
国家自然科学基金项目(No.41301371
41171289)
地理信息工程国家重点实验室开放基金项目(SKLGIE2014-M-3-3)
