基于人眼视觉特性的混合滤波算法

An mixed filter algorithm based on human visual system

为了有效地滤除混合噪声,本文提出了一种基于人眼视觉特性的混合滤波算法。该方法首先采用基于人眼视觉特性的噪声敏感系数作为阈值来确定脉冲噪声点,对检测出脉冲噪声点采用自适应窗口大小的迭代中值滤波进行滤波,而对于含有高斯噪声的像素点则采用一种保护细节的改进的自适应模糊滤波器进行处理。该算法与标准滤波方法及其它改进混合滤波算法相比,具有更好的滤波性能。

In order to remove mixed noise effectively, an mixed filter algorithm based on human vision system is proposed. First it determines possible impulse noisy pixels according to the noisy threshold, the noisy sensitive coefficient which is based on human vision system, then the impulse noisy pixels are moved by the iterative median filtering algorithm, and the filter window size is adaptively adjusted. A detail - preserving fuzzy filter is adopted to reduce Gaussian noise contained in other pixels. The algorithm has better filtering performance than the standard filters and other improved mixed filters.