提出了一种基于双密度双树复小波(double-density dual-tree complex wavelet transform, DDDT-CWT)基的结构化CS图像重构算法,该算法将图像在双密度双树复小波变换下的系数呈现的树结构化特征与CoSaMP重构算法相结合,实现了对原始图像的更精确重构.实验结果表明:在相同压缩比的前提下,与传统使用DWT基且未考虑变换系数结构化特征的重构算法相比,使用DDDT-CWT基和融入结构化特征的重构算法分别可获得2.9~3.2 dB与0.2~1.2 dB的增益,综合两者后的重构算法可获得3.8~4.3 dB以上的增益.
We propose a new structured compressed sensing recovery algorithm of images based on double-density dual-tree complex wavelet transform (DDDT-CWT). The algorithm combines the structured characteristic of coefficients under DDDT-CWT and compressive sample matching pursuit (CoSaMP) recovery algorithm. It has good reconstructed image performance. Simulation results show advantages of the proposed method as compared with traditional recovery algorithm using DWT basis and without considering structured characteristic of coefficients. With the same compression ratio, PSNR is improved by 2.9~3.2 dB and 0.2~1.2 dB when using the DDDT-CWT basis and considering structured characteristic respectively. The PSNR gain reaches 3.8~4.3 dB when combining these two features together.
[1] Donoho D L. Compressed sensing[J]. IEEE Transactions on Information Theory, 2006, 52(4):1289-1306.
[2] Colonnese S, Rinauro S, Cusani R, Scarano G. The restricted isometry property of the radon-like CS matrix[C]//2013 IEEE 15th International Workshop on Multimedia Signal Processing (MMSP), 2013:248-253.
[3] Hayashi K, Nagahara M, Tanaka T. A user's guide to compressed sensing for communications systems[J]. IEICE Transactions on Communications, 2013, 96(3):685-712.
[4] Davenport M A, Wakin M B. Analysis of orthogonal matching pursuit using the restricted isometry property[J]. IEEE Transactions on Information Theory, 2010, 56(9):4395-4401.
[5] Wang W, Ni L. Multipath subspace pursuit for compressive sensing signal reconstruction[C]//20147th International Congress on Image and Signal Processing (CISP), 2014:1141-1145.
[6] Sathyabama B, Sankari S S. Nayagara G S. Fusion of satellite images using compressive sampling matching pursuit (CoSaMP) method[C]//2013 Fourth National Conference on Computer vision, Pattern Recognition, Image Processing and Graphics (NCVPRIPG), 2013:1-4.
[7] Ekanadham C, Tranchina D, Simoncelli E P. Recovery of sparse translation-invariant signals with continuous basis pursuit[J]. IEEE Transactions on Signal Processing, 2011, 59(10):4735-4744.
[8] Li J, Shen Y, Wang Q. Stepwise suboptimal iterative hard thresholding algorithm for compressive sensing[C]//2012 IEEE International Instrumentation and Measurement Technology Conference (I2MTC), 2012:1332-1336.
[9] Lai L L, Wang Q P, Wang Q. Research on one kind of improved GPSR algorithm[C]//Computer Science and Electronics Engineering, 2012:715-718.
[10] 丰祥,万旺根. 运用压缩感知理论的图像稀疏表示与重建[J]. 应用科学学报,2014, 32(5):447-452. Feng X,Wan W G. Sparse representation and reconstruction of image based on compressed sensing[J] Journal of Applied Sciences, 2014, 32(5):447-452. (in Chinese)
[11] Duarte M, Davenport M, Takhar D. Single-pixel imaging via compressive sampling[J]. IEEE Signal Processing Magazine, 2008, 25(2):83-91.
[12] Goyal V K, Fletcher A K, Rangan S. Compressive sampling and lossy compression[J]. IEEE Signal Processing Magazine, 2008, 25(2):48-56.
[13] Selesnick I W, Baraniuk R G, Kingsbury N G. The dual tree complex wavelet transform[J]. IEEE Signal Processing Magazine, 2005, 22(6):123-151.
[14] Selesnick I W. The double-density dual-tree DWT[J]. IEEE Transactions on Signal Processing, 2004, 52(5):1304-1314.
[15] Baraniuk R G, Cevher V, Durate M F. Model-based compressive sensing[J]. IEEE Transactions on Information Theory, 2010, 56(4):1982-2001.
[16] La C, Do M N. Tree-based orthogonal matching pursuit algorithm for signal reconstruction[C]//IEEE International Conference on Image Processing (ICIP), 2006:1277-1280.
[17] Cen Y G, Wang F F, Zhao R Z, Cui L H, Cen L H, Miao Z J, Cen Y. Tree-based backtracking orthogonal matching pursuit for sparse signal reconstruction[J]. Journal of Applied Mathematics, 2013(8):1-8.
