期刊文献+

基于轮廓波与梯度组合稀疏的压缩感知重建 被引量:1

Compressive sensing reconstruction based on contourlet transform and gradient with sparse combination
下载PDF
导出
摘要 数字图像信号通常兼具平滑与边缘特征,从而导致单一稀疏空间的压缩感知效果信息损失大。而小波变换缺乏方向特征,常规轮廓波在频域上具有非局部化缺点,因而对边缘纹理信息刻画不够精确。针对该问题,一种迭代加权多稀疏空间组合重建算法得以提出。其同时采用对边缘特征描述更为高效的尖锐频率局部化轮廓波和对光滑背景表示更准确的离散梯度作为稀疏域,自适应选取权重因子来调整重建信号的稀疏特征,从而实现对信号特征更准确地刻画。实验表明,该算法在主观视觉效果和客观评价方面都远远优于其它组合稀疏算法。 Digital image signals usually have both smooth and edge features, lead that the recovery results of single-space compressive sensing have big information loss in most cases. On the other hand, wavelet transform lacks the directional characteristics and conventional contourlet transform has a non- localized disadvantage in the fl'equency domain, which give rise to an inefficiency in modeling the edge textures. In order to solve this problem, an iterative-weighted reconstruction algorithm with the combination of multiple sparse spaces was proposed, which involved two distinct sparse spaces into the reconstruction process: one is the sharp frequency localized contourlet transform space that can represent the edge features efficiently, and the other one is the gradient space that is more suitable to represent the smooth background regions. Besides, updating the weighting factor of each adaptively according to the characteristics of the signal itself, the algorithm can describe the features of the signal more accurately. The experimental results show that the proposed algorithm is much superior to other combined-space recovery algorithms both in subjective visual effects and objective assessments.
作者 李露 王良君 肖铁军 LI Lu;WANG Liang-jun;XIAO Tie-jun(School of Computer Science and Telecommunication Engineering,Jiangsu University,Zhenjiang 212013,Jiangsu Province,China)
出处 《信息技术》 2018年第10期28-33,39,共7页 Information Technology
基金 国家自然科学基金(61601202) 江苏省自然科学基金(BK20140571) 江苏大学高级专业人才科研启动基金(14JDG038)
关键词 图像重建 压缩感知 多稀疏空间 尖锐频率局部化轮廓波 梯度 image reconstruction compressive sensing muhiple sparse spaces sharp frequencylocalized contourlettransform gradient
  • 相关文献

参考文献6

二级参考文献136

  • 1D Donoho. Compressed sensing[ J]. IEEE Trans Inform Theory,2006,52(4) : 1289 - 1306.
  • 2M A T Figueiredo, R D Nowak, S J Wright. Gradient projection for sparse reconstruction: Appfication to compressed sensing and other inverse problems [ J ]. IEEE J Selected Topics in Signal Processing: Special Issue on Convex Optimization Methods for Signal Processing, 2007,1(4) :586 - 598.
  • 3I Daubechies, M Defrise, C De Mol. An iterative thresholding algorithm for finear inverse problems with a sparsity constraint [ J]. Comm Pure Appl Math,2004,57( 11 ):1413 - 1457.
  • 4T Blumensath, M Davies. Iterative hard thresholding for compressed sensing[ J]. Appl Comput Harmon Anal, 2009, 27 ( 3 ) : 265 - 274.
  • 5A C Gilbert, S Guha, P Indyk, S Muthukrishnan, M J Strauss. Near-optimal sparse Fourier representations via sampling[ A]. Proc. of the 2002 ACM Symposium on Theory of Computing STOC[C]. Montreal, Quebec, Canada, 2002. 152 - 161.
  • 6E Candbs, J Romberg, T Tao. Robust uncertainty principles: Exact signal reconstruction from highly incomplete frequency information [ J]. IEEE Trans Inform Theory ,2006,52(2) :489- 509.
  • 7E Candes, T Tao. Error correction via linear programming [A]. Proc. of 46th Annual IEEE Symposium on Foundations of Computer Science FOCS [ C ] . Pittsburgh, Pennsylvania, USA. 2005.295 - 308.
  • 8S Mallat, Z Zhang. Matching pursuit in a time-frequency dictionary[ J]. IEEE Trans Singal Processing, 1993,41 (12) : 3397 - 3415.
  • 9E J Candes, T Tao. Decoding by linear programming [ J ]. IEEE Trans Inform Theory,2005,51 (12):4203- 4215.
  • 10S J Kim, K Koh, M Lustig, S Boyd, D Gorinevsky. A interiorpoint method for large-scale ly-regularized least-squares problems with applications in signal processing and statistics[J]. Journal of Machine Learning Research, 2007,7 ( 8 ) : 1519 - 1555.

共引文献353

同被引文献18

引证文献1

二级引证文献3

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