摘要
针对同时感染脉冲噪声和高斯噪声的混合噪声图像,以全变分去噪模型为基础,结合中值滤波技术,提出一种新的混合噪声滤波算法。该算法首先根据脉冲噪声的特点和像素的局部能量信息,分离出脉冲噪声并以改进中值滤波算法去除,然后对含有高斯噪声的图像采用自适应广义变分模型进行降噪处理。实验结果表明,该算法在有效滤除混合噪声的同时能很好地保护图像细节,为去除图像中的混合噪声提供了一种有效的途径。
In light of the noise corrupted images affected by impulse noise and Gaussian noise mixed together,a new filtering algorithm for mixed noise is proposed based on total variational denoising model and combining median filtering technique.Firstly,the algorithm separates impulse noise from the mixed one according to its characteristic and local energy information of pixels,and then uses the improved median filter to remove it,afterwards the algorithm adopts the adaptive generalized variational model to perform denoising operation against the image with Gaussian noise.Simulation results show that the proposed algorithm can suppress mixed noise very effectively and preserve image details very well,this provides an effective means for removing mixed noise in corrupted images.
出处
《计算机应用与软件》
CSCD
2010年第10期98-102,共5页
Computer Applications and Software
基金
四川省教育厅自然科学重点基金(No.09ZA104)项目资助
关键词
混合噪声
噪声检测
中值滤波
全变分模型
Mixed noise
Noise detection
Median Altering
Total variational model
作者简介
王益艳，硕士，主研领域：图像处理，模式识别。
