自适应形态学滤波器研究

形态学滤波器是从数学形态学发展起来的一种新型非线性滤波器,由于其具有并行性与可快速实现的特点,在信号处理和图像分析领域中得到一定的应用,但形态学操作在抑制图像噪声的同时会模糊图像导致较多的细节损失,本文在现有的形态学滤波器基础上提出一种自适应的形态学滤波算法,此算法可以根据被处理图像局部特征的变化自适应选择最佳结构元进行形态学运算,通过实验和测试结果比较,其在噪声去除能力特别是局部细节保持方面都优于现有的形态学滤波器。

An Adaptive Morphological Filter Preserving Partial Details of Images

The morphological filter is a new kind of nonlinear filter,which has been developed from mathematic morphology and well applied in signal processing and image analysis,but the operations of morphological filter also blur images and destroy some important partial details.Based on the fact,a new adaptive algorithm of morphological filter is proposed in this paper,the proposed algorithm adaptively chooses the best structuring element according to the characteristic of images partial details,and utilizes morphological operations to remove noises.Theoretical and experimental results show the superiority of this approach including high ability of removing noises and preserving the important partial details of images in comparison with some recent methods.