一种图像自适应中值滤波算法

An Image Adaptive Median Filter Algorithm

图像在产生、传递与接收过程中,会受到外界的影响尤其是噪音的干扰,严重影响图像质量,有必要进行预处理以保证图像的正常使用。阐述了脉冲噪音理论,分析了一维与二维中值滤波原理,提出一种自适应中值滤波算法。该算法作为非线性滤波算法,弥补了线性滤波算法缺陷,明显抑制了噪音,保证了原图像的信息内容。实验结果表明,这种自适应中值滤波算法在归一化均方误差(NMSE)、峰值信噪比(PSNR)、算法耗时T(s)等方面,比中值滤波算法具有一定的优势。

In the process of image creation,transmission and reception,there will be an adverse impact on the outside situation,especially the noise interference which will seriously affect the quality of the image,and result in blur and distortion of the consequences.It is necessary to have image pre-processing to ensure its normal use.Based on the pulse noise theory,we take sequence analysis of 1 Dand 2 Dmedian filtering principle,and then propose an adaptive median filtering algorithm.As a nonlinear filtering algorithm.This algorithm compensates for the deficiency of linear filtering algorithm,restrain the noise,and ensures the information content of the original image.The experimental results show that this adaptive median filtering algorithm has some advantages compared to median filtering algorithm on experimental data for Normalized Mean Squared Error（NMSE）,Peak Signal-to-Noise Ratio（PSNR）and T（s）three aspects.