摘要
以层叠滤波器理论为基础,通过对信号进行阈值分解和引入全方位结构元,提出了一种用于处理噪声图象的多级全方位细节保持层叠滤波器。通过在滤波算法中引入细节检测单元,使得滤波器的第2级可以针对噪声污染图象的不同区域,自动选择相应的滤波器单元,从而达到滤除噪声、保持图象边缘细节的效果.仿真实验验证了这种滤波器的有效性。
Based on stack filter theory, a new filter called multi-level omnidrectional detail preserving stack filter is developed for processing noisy image, by threshold decomposition of input signal and using multi-directional structural elements. The second stage of the filter can automatically select corresponding positive Boolean function (PBF) according to the different area of input noisy image by designing detail detection unit in the filtering algorithm. So the class of filters can effectively preserve imag...
出处
《吉林大学学报(信息科学版)》
CAS
2003年第S1期21-24,共4页
Journal of Jilin University(Information Science Edition)
关键词
非线性滤波
层叠滤波器
阈值分解
正布尔函数
Nonlinear filtering
Stack filter
Threshold decomposition
Positive boolean function