含缺失成分的矩阵的广义低秩逼近及其在图像处理中的应用.pdfVIP

第27 卷 第 11 期 计算机辅助设计与图形学学报 Vol. 27 No.11 2015 年 11 月 Journal of Computer-Aided Design Computer Graphics Nov. 2015 含缺失成分的矩阵的广义低秩逼近及其在图像处理中的应用 1) 2) 1) 李 璐 , 董秋雷 , 赵瑞珍 1) (北京交通大学计算机与信息技术学院信息科学研究所 北京 100044) 2) ( 中国科学院自动化研究所模式识别国家重点实验室 北京 100190) ) 摘 要: 针对在许多实际应用中数据以矩阵形式而非向量形式存在的问题, 重点讨论含缺失成分的矩阵低秩逼近问 题的广义版本, 即如何对一组含缺失成分的矩阵进行低秩逼近. 首先构造一个最优化问题来表达原始的广义低秩逼 近问题, 该最优化问题最小化输入矩阵组中已知成分的总重构误差; 然后提出了一种迭代优化算法来求解上述的最 优化问题; 最后给出详细的算法分析. 大量的模拟实验与真实图像实验结果表明, 文中算法具有较好的性能. 关键词：广义低秩逼近; 缺失成分; 重构误差; 迭代优化算法 中图法分类号：TP391 Generalized Low-rank Approximations of Matrices with Missing Components and its Applications in Image Processing Li Lu1), Dong Qiulei2), and Zhao Ruizhen1) 1) (Institute of Information Science, School of Computer and Information Technology, Beijing Jiaotong University, Beijing 100044) 2) (National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Sciences, Beijing 100190) Abstract: Considering that data used in many applications are intrinsically in matrix form rather than in vector form, this paper focuses on the generalized version of the problem of a low-rank approximation of a matrix with missing components, i.e. low-rank approximations of a set of matrices with missing components. This generalized problem is formulated as an optimization problem at first, which minimizes the total reconstruction error of the known components in these matrices. Then, an iterative algorithm is designed for calculating the generalized low-rank approximations of matrices with missing components, called GLRAMMC. Finally, detailed algorithmic analysis is given. Extensive experimental results on synthetic data as well as on real image data show the ef