期刊文献+

快速图像调和稀疏分解模型及其应用

Fast Harmonic and Sparse Image Decomposition Model and Its Application
下载PDF
导出
摘要 首先提出一种图像调和稀疏分解(HSID)模型,用于将一幅图像分解为调和分量和稀疏分量.然后提出基于增广拉格朗日交替方向法(ALADM)的HSID求解算法(HSID_ALADM),算法每次迭代的主要计算量为矩阵的快速傅氏变换,因此HSID_ALADM快速高效.将HSID_ALADM用于红外图像分解,所得的调和分量可视为图像背景,而其稀疏分量可视为图像中的目标分量,通过搜索稀疏分量中的局部能量极值,可检测出红外图像中的小目标.HSID_ALADM亦可直接用于图像补全与修复.实际的红外图像目标检测及图像补全与修复实验表明HSID_ALADM性能良好. An image decomposition model, harmonic and sparse image decomposition ( HSID), is firstly put forward to decompose an image into a harmonic component and a sparse component. Then, based on augmented Lagrangian alternating direction method (ALADM), an algorithm, namely HSID_ALADM, is presented to solve HSID. The main computational load of each iteration in HSID ALADM is computing fast Fourier transform (FFT), which makes HSID _ALADM fast. HSID _ALADM can be used to decompose an infrared image with small targets into a harmonic component and a sparse component. The harmonic component is considered as the modeling of the background, and the sparse component as the small target component. By searching for the maximum local energy regions in the sparse component, the infrared targets in the infrared image can be easily and accurately located. Experimental results of small infrared target detection for real infrared images and image completion and inpainting show good performance of HSID_ALAD.
作者 郑成勇
出处 《模式识别与人工智能》 EI CSCD 北大核心 2014年第6期546-553,共8页 Pattern Recognition and Artificial Intelligence
基金 国家自然科学基金项目(No.61075116) 五邑大学青年科研基金项目(No.2013zk15)资助
关键词 图像分解 增广拉格朗日乘子 交替方向法 红外目标检测 图像修复 Image Decomposition, Augmented Lag-range Multiplier, Alternating Direction Method,Infrared Target Detection, Image Inpainting
  • 相关文献

参考文献13

  • 1Wright J, Peng Y G, Ma Y, et al. Robust Principal ComponentAnalysis : Exact Recovery of Corrupted Low-Rank Matrices by Con-vex Optimization[ EB/OL]. [2013-04-10]. http://perception,csl. illinois. edu/matrix-rank/Files/nips2009. pdf.
  • 2Lin Z C, Liu R S, Su Z X, et al. Linearized Alternating DirectionMethod with Adaptive Penalty for Low Rank Representation //Shawe-Taylor J,Zemel R S,Bartlett P,et al.,eds. Advances inNeural Information Processing Systems. Granada, Spain : Springer,2011, 24: 612-620.
  • 3Cand^s E J, Li X D, Ma Y, et al. Robust Principal ComponentAnalysis. Journal of the ACM, 2011, 58(3) : 1 -39.
  • 4Tan W T,Cheung G, Ma Y. Face Recovery in Conference VideoStreaming Using Robust Principal Component Analysis // Proc of theInternational Conference on Image Processing. Brussels, Belgium,2011 : 3225-3228.
  • 5Cand6s E J, Recht B. Exact Matrix Completion via Convex Optimi-zation. Foundations of Computational Mathematics, 2009 , 9(6);717-772.
  • 6Toh K C, Yun S W. An Accelerated Proximal Gradient Algorithmfor Nuclear Norm Regularized Linear Least Squares Problems. Pacif-ic Journal of Optimization, 2010, 6(3) : 615-640.
  • 7Cai J F, Cand^s E J, Shen Z W. A Singular Value Thresholding Al-gorithm for Matrix Completion. SIAM Journal on Optimization,2010,20(4): 1956-1982.
  • 8Ma S Q,Goldfarb D, Chen L F. Fixed Point and Bregman IterativeMethods for Matrix Rank Minimization. Mathematical Programming :Series A, 2011, 128(1/2) : 321-353.
  • 9Chan T F,Shen J H,Vese L. Variational PDE Models in ImageProcessing. Notices of the American Mathematical Society, 2003 ,50(1): 14-26.
  • 10Gabay D, Mercier B. A Dual Algorithm for the Solution of Nonlin-ear Variational Problems via Finite Element Approximations. Com-puters and Mathematics with Applications, 1976, 2(1): 17-40.

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部