基于灰色关联度的图像自适应中值滤波算法

An Adaptive Algorithm for Median Filter of Image Based on Grey Relation

研究图像去噪是图像处理中一个重要的课题。传统去除高斯噪声的方法是均值滤波,均值滤波方法模糊了图像的细节信息,中值滤波虽然能较好的保护图像的细节信息但滤除高斯噪声的效果不理想。灰色关联度能根据系统各因素间相似程度来度量因素之间的关联程度。利用灰色关联度的特性和中值滤波器的优点,提出一种基于灰色关联度的自适应中值滤波算法,算法可以根据图像像素之间的相似程度自适应地调整滤波加权系数,使滤波系数更加合理,改善原有算法的滤波性能。实验表明算法对受到高斯噪声干扰的图像进行去噪取得较好的滤波效果,同时还保护了图像的细节信息。

Removing noise from the original image is an important problem for researchers.Traditional method for removing Gaussian noise is Mean filter,but it blurs the details of the image.Another traditional method is the Median filter,it can protect the edge of the image,but at the same time,the result is worse than expected.Grey relation can measure the degree of correlation among the levels of factor based on the factors similar to the system.Because of the Characteristics of the grey relation and the advantage of the median filter,an efficient algorithm based on grey relation and median filter to remove the noise in the image is proposed.This algorithm can adjust the filter coefficients adaptively,according to various pieces of the image features and it can improve the validity of the previous algorithms.Experiment shows that the algorithm can get good effect from the image interfered by Gaussian noise resource and also retain details well.