|本期目录/Table of Contents|

[1]韦 仙,康睿丹.基于降维压缩法的图像重构[J].武汉工程大学学报,2015,37(12):69-74.[doi:10. 3969/j. issn. 1674-2869. 2015. 12. 015]
 -.Image reconstruction based on dimension reduction and compression technology[J].Journal of Wuhan Institute of Technology,2015,37(12):69-74.[doi:10. 3969/j. issn. 1674-2869. 2015. 12. 015]
点击复制

基于降维压缩法的图像重构(/HTML)
分享到:

《武汉工程大学学报》[ISSN:1674-2869/CN:42-1779/TQ]

卷:
37
期数:
2015年12期
页码:
69-74
栏目:
机电与信息工程
出版日期:
2016-01-14

文章信息/Info

Title:
Image reconstruction based on dimension reduction and compression technology
文章编号:
1674-2869(2015)12-0069-06
作者:
韦 仙康睿丹
太原工业学院理学系,山西 太原 030008
Author(s):
WEI Xian KANG Rui-dan
Faculty of Science, Taiyuan Institute of Technology, Shanxi 030008, China
关键词:
矩阵填充人脸识别低秩奇异值分解
Keywords:
matrix completion face recognition low?鄄rank singular value decomposition
分类号:
O411.1
DOI:
10. 3969/j. issn. 1674-2869. 2015. 12. 015
文献标志码:
A
摘要:
针对人脸图像易受环境因素的影响造成缺失或者受噪声污染,提出了从有限的信息中重构完整的图像矩阵的方法. 首先利用奇异值压缩降维的方法提取人脸图像的特征值,并运用基于凸优化的矩阵填充技术对缺失的图像矩阵进行有效重构,然后采用固定点迭代算法,通过Matlab语言编程,进行分裂法迭代,在选取合适参数的情况下使运行程序快速收敛至目标矩阵,减小了运行时间. 分析峰值信噪比随奇异值个数的变化关系,对人脸图像的保真度进行评估,通过对不同采样率下人脸图像重构效果的对比,运行时间的分析,得出降维压缩技术能够有效实现图像矩阵填充的结论.
Abstract:
Aimed at that the face image is usually missing and corrupted by noise under the impact of environmental factors, we proposed a method to reconstruct the complete image matrix from the limited information. Firstly, we applied the matrix completion theory to reconstruct the image matrix whose eigenvalues are effectively extracted using the method of singular value compression. Then, we used the matrix completion technology based on the convex optimization to study the problem of missing matrix reconstruction by running the fixed point iterative algorithm. This algorithm can quickly converge to the target matrix in the case of selecting appropriate parameters by conducting splitting iteration with the help of Matlab programming language,which reduces the running time. We evaluated the fidelity of the face image by analyzing the relationship between the peak signal to noise ratio and the number of singular values. The conclusion shows that the image matrix is effectively completed using the technology of dimension compression through analyzing the effect of face image reconstruction under different sampling rates and the run-times.

参考文献/References:

[1] 杨济美,向世明,刘荣,等.矩阵低秩逼近的快速增量算法及其在人脸图像中的应用[J].中国科学技术大学学报,2009,39(9):970-979.YANG Ji-mei, XIANG Shi-ming, LIU Rong, et al. A fast incremental algorithm for low rank approximations of matrices and its applications in facial images[J]. Journal of University of Science and Technology of China, 2009,39(9):970-979.(in Chinese)[2] HONG Z Q. Algebraic feature extraction of image for recognition[J].Pattern Recognition,1991,24(3):211?鄄219.[3] BEGHDADI A,PESQUET P B. A new image distortion measure based on wavelet decomposition[J].Proc of IEEE ISSPA,2003(1):485?鄄488.[4] 夏平平,吕太之. 动态人脸识别系统的设计与实现[J].武汉工程大学学报,2011,33(10):107?鄄110.XIA Ping?鄄ping, LYU Tai?鄄zhi. Design and implementation of a dynamic faces recognition system[J]. Journal of Wuhan Institute of Technology, 2011,33(10):107?鄄110.(in Chinese)[5] 韦仙. 基于矩阵填充技术重构哦低秩密度矩阵[J].武汉工程大学学报,2015,37(2):72?鄄76.WEI Xian. Reconstructing low?鄄rank density matrix via matrix completion[J]. Journal of Wuhan Institute of Technology, 2015,37(2):72?鄄76. (in Chinese)[6] EMMANUEL J C,BENJAMIN R. Exact low?鄄rank matrix completion via convex optimization[J]. IEEE,2008(23?鄄26): 806?鄄812.[7] MA Shi?鄄qian,DONALD G, CHEN Li?鄄feng. Fixed point and Bregman iterative methods for matrix rank minimization[J]. Mathematical Programming,2011,128(1?鄄2):321?鄄353.

相似文献/References:

[1]夏平平,吕太之.动态人脸识别系统的设计与实现[J].武汉工程大学学报,2011,(10):107.
 XIA Ping ping,LV Tai zhi.Design and implementation of dynamic faces recognition system[J].Journal of Wuhan Institute of Technology,2011,(12):107.
[2]韦 仙.基于矩阵填充技术重构低秩密度矩阵[J].武汉工程大学学报,2015,37(02):72.[doi:10. 3969/j. issn. 1674-2869. 2015. 02. 016]
 Reconstructing low-rank density matrix via matrix completion[J].Journal of Wuhan Institute of Technology,2015,37(12):72.[doi:10. 3969/j. issn. 1674-2869. 2015. 02. 016]
[3]陈希彤,卢 涛*.基于全局深度分离卷积残差网络的高效人脸识别算法[J].武汉工程大学学报,2019,(03):276.[doi:10. 3969/j. issn. 1674-2869. 2019. 03. 014]
 CHEN Xitong,LU Tao *.Efficient Face Recognition Algorithm Using Global Deep Separable Convolutional and Residual Network[J].Journal of Wuhan Institute of Technology,2019,(12):276.[doi:10. 3969/j. issn. 1674-2869. 2019. 03. 014]
[4]郭 婷,张天序*,郭诗嘉.一种红外图像和宽光谱融合的人脸识别算法[J].武汉工程大学学报,2022,44(03):320.[doi:10.19843/j.cnki.CN42-1779/TQ.202111015]
 GUO Ting,ZHANG Tianxu*,GUO Shijia.Face Recognition Algorithm Based on Fusion of Infrared Image and Wide Spectrum Features[J].Journal of Wuhan Institute of Technology,2022,44(12):320.[doi:10.19843/j.cnki.CN42-1779/TQ.202111015]

备注/Memo

备注/Memo:
收稿日期:2015-11-05基金项目:太原工业学院院级青年科学基金(2014LQ05)作者简介:韦仙(1988-),女,山西晋城人,助理实验师,硕士. 研究方向:压缩感知与矩阵填充.
更新日期/Last Update: 2016-01-21