新融合小波变换与低秩矩阵恢复图像去噪算法  

New image denoising algorithm based on wavelet transform and low rank matrix recovery

作　　者：吴君钦[1] 邬亮[1] 

机构地区：[1]江西理工大学信息工程学院

出　　处：《电视技术》2016年第3期17-21,共5页Video Engineering

基　　金：国家自然科学基金项目(11062002)

摘　　要：针对小波阈值图像去噪会引入量化噪声和阈值选取不当会损失图像本身有用信息的问题,提出一种新的融合小波变换与低秩矩阵恢复(Low Rank Matrtix Recovery,LRMR)的图像去噪算法。不同于传统的单一阈值的去噪算法,所提出的算法在单一阈值上结合了低秩矩阵恢复算法,这样不仅能进一步消除噪声,同时还能修复被噪声损坏的数据,而且更能适应各种不同的噪声环境。首先,选取固定阈值对图像矩阵进行小波去噪处理。其次,采用增广拉格朗日乘子算法最小化矩阵核范数。最后,将矩阵分解为低秩逼近矩阵和稀疏误差矩阵。实验结果表明,算法获得了较高的峰值信噪比,在不同噪声环境下有较高的鲁棒性。Aiming at the problem of image denoising based on the wavelet threshold introduced the quantization noise, the threshold selected may lose the useful information of the image itself. An image denoising algorithm based on wavelet transform and low rank matrix recovery （LRMR） is proposed. Different from a single threshold denoising algorithm, the algorithm can eliminate the more noise and fixs the corrupted data, and can adapt to various noise environment. Firstly, select the fixed threshold to carry on the wavelet denoising to preproeess the image matrix. Then the augmented Lagrange multiplier algorithm minimize the kernel matrix norm, and the preproeessed matrix factorized into a low rank approximation matrix and a sparse error matrix. Experiments results show that the proposed algorithm has better performance and robustness under different noise condition.

关 键 词：小波变换 图像去噪 增广拉格朗日乘子 低秩矩阵恢复 最小核范数 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]