广东工业大学学报2017，Vol.34Issue(6)：43-48,6.DOI:10.12052/gdutxb.170006
基于L1/2自适应稀疏正则化的图像重建算法
A Super-resolution Image Reconstruction Algorithm with Adaptive L1/2 Sparse Regularization
摘要
Abstract
In order to solve the ill-posing problem and poor effect of fixed regularization parameter in superresolution image reconstruction,an adaptive regularization combining the study of sparse representation is proposed.By additional restrictions for compatibility of adjacent patches,a new L1/2 non-convex optimization model is built.Reweighted L2 Norm rather than Lp (0＜p＜l) Norm is applied into the adaptive algorithm for adjustment of regularization parameter.With the help of joint dictionary training method,some important features for improving the quality of reconstructed image are obtained.Experimental results show that the method has significant advantages in denoising and preserving edge details.It is showed that the proposed method not only makes the desired high-resolution images visually clearer,but it also outperforms some traditional methods in both the value of peak signal to noise ratio and structural similarity.关键词
L1/2非凸优化/稀疏表示/自适应正则化/超分辨率重建/邻近块兼容性/拼接字典Key words
L1/2 non-convex optimization/sparse representation/adaptive regularization/super-resolution reconstruction/compatibility of adjacent patches/jointing dictionary分类
信息技术与安全科学引用本文复制引用
叶向荣,刘怡俊,陈云华,熊炯涛..基于L1/2自适应稀疏正则化的图像重建算法[J].广东工业大学学报,2017,34(6):43-48,6.基金项目
广东省自然科学基金资助项目(2014A030310169,2016A030313713) （2014A030310169,2016A030313713）
广东省科技计划项目(2016B090918126,2016B090904001,2014B090901061,2015B090901060,2015B090908001,2015B090903080) （2016B090918126,2016B090904001,2014B090901061,2015B090901060,2015B090908001,2015B090903080）
广州市科技计划项目(2014Y2-00211) （2014Y2-00211）
