基于非噪声像素重构的PK-SVD脉冲噪声滤波  

PK-SVD Filter for Impulse Noise Based on Non-noisy Pixel Reconstruction

作　　者：黄宴委[1] 祁冰露[1] 

机构地区：[1]福州大学电气工程与自动化学院,福州350108

出　　处：《模式识别与人工智能》2014年第11期977-984,共8页Pattern Recognition and Artificial Intelligence

基　　金：国家自然科学基金项目(No.61174051);高校博士点新教师基金项目(No.20113514120007)资助

摘　　要：提出一种基于非噪声像素重构的K-SVD(Pixel K-SVD)脉冲噪声滤波方法.在图像重构阶段,以非噪声点像素值为优化目标,利用分层重构改进OMP算法求解优化函数,获得重构图像以提高恢复图像质量;在字典训练阶段,PK-SVD不再固定原子的系数,而是使用重复奇异值分解同时更新原子和系数.将PK-SVD与其他3种方法进行比较,实验结果表明,PK-SVD能得到最稀疏化的字典,较好地抑制脉冲噪声,使得滤波图像较清晰且具有较高的峰值信噪比.An improved K-SVD method based on non-noisy impulse noise. In the phase of image reconstruction, pixel reconstruction （PK-SVD） is proposed to fiher non-noisy pixels are applied in the construction of optimal function to obtain the reconstructed image and improve the filtering performance, and the optimal function is solved by integrating the hierarchical property into the OMP algorithm. In the phase of dictionary training, PK-SVD uses the iterant K-singular value decomposition to renovate both atoms and their coefficients rather than fixes the coefficients. The simulation results show that compared with the other three methods, PK-SVD obtains the sparsest dictionary and the clearest image with higher peak signal to noise ratio.

关 键 词：脉冲噪声滤波 非噪声像素重构 K-SVD 分层OMP 字典训练 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]