一种基于Stein无偏风险估计的复合去噪算法

发布时间：2017-12-27 08:26

本文关键词：一种基于Stein无偏风险估计的复合去噪算法　出处：《电讯技术》2016年11期 　论文类型：期刊论文

【摘要】：针对统计非局部均值滤波损坏图像的细节与鲁棒性双边带滤波去噪不充分的缺点,提出了一种基于统计非局部均值滤波与鲁棒性双边带滤波相结合的复合滤波算法。该复合滤波算法通过统计非局部均值滤波与鲁棒性双边带滤波线性组合,利用Stein无偏风险估计对复合算法中的参数进行估计。实验中,从主观与客观方面进行对比分析,证明所提出的复合算法体现了非局部均值滤波与双边带滤波的优点,能有效地去除噪声并更好地保留图像的细节信息,峰值信噪比提高1~2 d B。
[Abstract]:Aiming at the shortcomings of the statistical nonlocal mean filter, which is the details and robustness of image denoising and the side-band filtering denoising is not enough, a compound filtering algorithm based on statistical nonlocal mean filter and robust bilateral band filtering is proposed. The composite filtering algorithm is composed of statistical nonlocal mean filter and robust bilateral band filtering. Stein's unbiased risk estimation is used to estimate the parameters in the composite algorithm. From the subjective and objective aspects, it is proved that the proposed composite algorithm embodies the advantages of non local mean filter and bilateral band filtering, which can effectively remove noise and preserve the details of the image better, and the peak signal-to-noise ratio is increased by 1~2 D B.
【作者单位】：四川理工学院自动化与电子信息学院;中国西南电子技术研究所;
【基金】：四川省教育厅项目(14ZB0211;14ZA0202) 人工智能四川省重点实验室开放基金项目(2015RZY01)~~
【分类号】：TP391.41
【正文快照】： 1引言非局部均值[1-2](Non-Local Means,NLM)与双边带滤波[3-4](Bilateral Filter,BF)是近几年兴起图像去噪技术。非局部均值去噪的思想是图像受干扰点的像素值可以通过图像其他区域相似块加权平均而得到,其去噪性能远好于传统均值滤波。但非局部均值仍存在不足:由于非局部均

【相似文献】