摘要
针对图像压缩感知(CIS)组稀疏表示重构算法在低采样率下尤其是对纹理特征相对复杂的图像重构质量不佳的问题,提出基于残差补偿的组稀疏表示(RCGSR)重构方法。对稀疏处理前后两幅图像对应位置图像组稀疏系数的残差进行稀疏化处理并补偿至后者的图像组稀疏系数中。归纳一种自适应软阈值收缩方案,对不同稀疏系数残差采取不同的阈值进行收缩处理,增强算法的鲁棒性。仿真结果表明,与目前性能最好的图像压缩感知重构算法GSR相比,所提算法在低采样率时显著提高了图像的重构性能。
To solve the problem of the lower reconstruction quality of group sparse representation of compressed image sensing(CIS)under low sampling rate especially the complex texture feature,agroup sparse representation recovery algorithm based on residual compensation(RCGSR)was proposed.The residual coefficients between corresponding patch groups of two images before and after sparsely represented were sparsified,and the patch group sparse coefficient of the latter for these residual was compensated.To enhance the robustness of the algorithm,an adaptive soft-threshold shrinkage scheme was deduced to deal with different sparse coefficient residual disposed by different threshold shrinkage.Simulation results show that compared with the best image compressed sensing reconstruction algorithm called GSR,the proposed algorithm significantly improves the performance at low sampling rate.
引文
[1]Hou B,Zhang G,Li Z,et al.Sparse coding-inspired highresolution ISAR imaging using multistage compressive sensing[J].IEEE Transactions on Aerospace and Electronic Systems,2017,53(1):26-40.
[2]Li S,Da Xu L,Wang X.Compressed sensing signal and data acquisition in wireless sensor networks and internet of things[J].IEEE Transactions on Industrial Informatics,2013,9(4):2177-2186.
[3]Bostock MJ,Holland DJ,Nietlispach D.Compressed Sensing1-norm minimisation in multidimensional NMR spectroscopy[M]//Fast NMR Data Acquisition,2017:267-303.
[4]Huang Y,Paisley J,Lin Q,et al.Bayesian nonparametric dictionary learning for compressed sensing MRI[J].IEEE Transactions on Image Processing,2014,23(12):5007-5019.
[5]Li S,Da Xu L,Wang X.Compressed sensing signal and data acquisition in wireless sensor networks and internet of things[J].IEEE Transactions on Industrial Informatics,2013,9(4):2177-2186.
[6]SHEN Yanfei,LI Jintao,ZHU Zhenmin,et al.Image reconstruction algorithm of compressed sensing based on nonlocal similarity model[J].Acta Automatica Sinica,2015,41(2):261-272(in Chinese).[沈燕飞,李锦涛,朱珍民,等.基于非局部相似模型的压缩感知图像恢复算法[J].自动化学报,2015,41(2):261-272.]
[7]SONG Yun,LI Xueyu,SHEN Yanfei,et al.Compressed sensing image reconstruction based on low rank of non-local similar patches[J].Acta Electronica Sinica,2017,45(3):695-703(in Chinese).[宋云,李雪玉,沈燕飞,等.基于非局部相似块低秩的压缩感知图像重建算法[J].电子学报,2017,45(3):695-703.]
[8]Zhang J,Zhao D,Zhao C,et al.Image compressive sensing recovery via collaborative sparsity[J].IEEE Journal on Emerging and Selected Topics in Circuits and Systems,2012,2(3):380-391.
[9]Zhang J,Zhao C,Zhao D,et al.Image compressive sensing recovery using adaptively learned sparsif-ying basis via L0minimization[J].Signal Processing,2014,103(10):114-126.
[10]Zhang J,Zhao D,Gao W.Group-based sparse representation for image restoration[J].IEEE Transactions on Image Processing,2014,23(8):3336-3351.
[11]Kumar BKS.Image denoising based on non-local means filter and its method noise thresholding[J].Signal,Image and Video Processing,2013,7(6):1211-1227.
[12]Dong W,Shi G,Ma Y,et al.Image restoration via simultaneous sparse coding:Where structured sparsity meets Gaussian scale mixture[J].International Journal of Computer Vision,2015,114(2-3):217-232.