摘要
现有的二维指数熵阈值分割快速算法的计算效率或收敛精度尚不够高,为此,本文提出了基于Tent映射混沌粒子群的二维直方图斜分指数熵阈值选取方法。首先引入了直方图区域斜分方法以改善分割结果的准确性和抗噪性,然后提出利用基于Tent映射混沌粒子群算法寻找最佳分割阈值,提高搜索过程的收敛精度和计算效率。实验结果表明:与基于灰度级-平均灰度级直方图直分的快速算法相比,该方法由于尽可能地考虑了所有目标点和背景点,分割效果更佳,同时以混沌粒子群优化搜索过程,运行时间更少;与基于灰度级-梯度直方图及Logistic混沌粒子群的方法相比,本文方法的抗噪性能更稳健、收敛精度更高。
In view of the inefficient shortage of computational efficiency or convergence precision of the existing thresholding methods based on 2-D exponent entropy, an improved 2-D oblique exponent entropy method based on tent map chaotic particle swarm algorithm is proposed in this paper. To achieve higher segmented accuracy and stronger anti-noise, an oblique regional division mode for histogram based thresholding method is introduced firstly, and then the chaotic particle swarm algorithm based on improved tent map is used to search for the optimal threshold so as to improve the convergence precision and computational efficiency. Compared with the fast thresholding method based on gray scale-average gray scale histogram of vertical regional division, experimental results show that the new method achieves a better segmentation quality since the entire object and background inner points are considered. And because the searching process of the new method is optimized by using chaotic particle swarm algorithm, the running time is reduced. Compared with the method based on gray scale-gradient histogram and logistic map chaotic particle swarm algorithm, a stronger anti-noise performance and higher convergence precision are obtained.
出处
《信号处理》
CSCD
北大核心
2010年第5期703-708,共6页
Journal of Signal Processing
基金
国家自然科学基金项目(60872065)
