摘要
针对三维Renyi熵阈值算法的高计算复杂性及抗噪性差的问题,提出一种基于三维直方图重建和降维的Renyi熵阈值分割算法。该方法通过重建三维直方图,将三维直方图区域划分由八分法改为二分法,不仅减弱了噪声干扰,而且计算复杂度也由O(L6)降到O(L)。实验结果表明,该算法抗噪性更强,分割效果更好,计算时间也远低于传统三维Renyi熵算法。
According to the high computational complexity and poor anti-noise performance of classical 3D Renyi entropical thresholding,this paper proposed a new 3D Renyi entropy image thresholding method.This method rebuilt the 3D histogram and transfered the region in 3D histogram from eight partitions into two partitions,which not only reduced the noise interfe-rence but also lowered the computational complexity from O（L6） to O（L）.The results show that the presented method has better anti-noise performance and visual quality,and lower time complexity compared with original 3D Renyi entropy＇s method.
出处
《计算机应用研究》
CSCD
北大核心
2013年第4期1223-1225,共3页
Application Research of Computers
基金
中央高校基本科研业务费基金资助项目(CDJXS11100032)
