论文部分内容阅读
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。
The maximum inter-class variance threshold method based on 3D histogram (3D Otsu) considers the neighborhood mean and median information, and has good anti-noise performance and satisfactory segmentation results. However, its computational complexity is very high and its efficiency is low. Firefly Algorithm (FA) is a new type of heuristic algorithm. Based on the introduction of the basic principle of firefly algorithm, this paper proposes a clustered firefly algorithm (CBLFA) based on Levi Flight, which is used to improve the efficiency of the three-dimensional Otsu threshold method. Experimental results show that this method can quickly obtain the appropriate threshold. The fitness function is generally better than the basic firefly algorithm and the basic particle swarm optimization algorithm. It is a robust three-dimensional Otsu threshold segmentation method.