Multilevel minimum cross entropy threshold selection based on honey bee mating optimization
Multilevel minimum cross entropy threshold selection based on honey bee mating optimization
复制标题
DOI：
--
复制
发表时间：
2009-01
期刊：
影响因子：
--
通讯作者：
M. Horng
中科院分区：
文献类型：
--
作者：
M. Horng
图像熵阈值分割方法在图像分割中引起了广泛的关注。本文主要研究了基于最小交叉熵准则的多级阈值分割问题。在文献中，采用粒子群算法进行阈值选择。本文采用的算法是蜜蜂交配优化算法（HBMO）。在实验中，采用两种不同的方法与分割结果进行比较。与粒子群算法的结果相比，HBMO的阈值选择更接近穷举搜索方法的最优值。HBMO的分割效果优于粒子群算法，但仍比粒子群算法慢。
Image entropy thresholding approach has drawn the attentions in image segmatation. The endeavor of this paper is focused onmultilevel thresholding using the minimumcross enrtop criterion. In the literature, the particle swarm optimization (PSO) had been applied to conducting the thresold selection. The adopted algorithm used in this paper is the honey bee mating optimization (HBMO). In experimnats, the two different methods are implemented for comparison with the results of segmantation. Compared to the results of PSO, the threshold selection of HBMO is more close to the optimal ones examined by the exhaustive search method. Furthermore, the segmentation result of HBMO is superior to PSO method, but it still slower than ones of PSO.