基于多字典和稀疏噪声编码的图像超分辨率重建
|
摘要
稀疏表示模型是通过将字典中的原子进行组合得到期望的结果.为了解决传统字典学习中所有图像块重建均使用同一个字典,从而忽略了最佳稀疏域的问题,提出来一种基于多字典和稀疏噪声编码的图像超分辨率重建算法.在字典训练时,利用图像的特征将它们合理地划分成若干个簇,每个聚类训练生成子字典对,利用最佳字典对进行重建.在求解稀疏系数阶段,引入稀疏编码噪声去除噪声的影响,利用图像非局部自相似性来获得原始图像稀疏编码系数的良好估计,然后将观测图像的稀疏编码系数集中到这些估计当中.实验表明,与ASDS算法和SSIM算法相比较,该算法有更好的重建结果,获得了更丰富的图像细节和更清晰的边缘.
The sparse representation model can obtain the desired result by combining the atoms in the dictionary. In order to solve the problem that all image blocks are reconstructed by using the same dictionary in the traditional dictionary learning,and the optimal sparse domain is ignored,an image reconstruction algorithm based on multi-dictionary and sparse-noise coding is proposed. The image blocks are reasonably divided into several clusters according to the features of the image in dictionary training. Each cluster is trained to generate sub-dictionary pairs,and the best dictionary pairs are used for reconstruction. In the stage of solving the sparse coefficient,the effect of noise is removed by introducing sparse coding noise,a good estimation of the sparse-coding coefficient of the original image is obtained by using the non-local self similarity of the image,and then the sparse-coding coefficients of the observed image are concentrated into these estimates. Experiments show that compared with ASDS algorithm and SSM algorithm,this algorithm has better reconstruction results,and gets richer image details and clearer edges.
The sparse representation model can obtain the desired result by combining the atoms in the dictionary. In order to solve the problem that all image blocks are reconstructed by using the same dictionary in the traditional dictionary learning,and the optimal sparse domain is ignored,an image reconstruction algorithm based on multi-dictionary and sparse-noise coding is proposed. The image blocks are reasonably divided into several clusters according to the features of the image in dictionary training. Each cluster is trained to generate sub-dictionary pairs,and the best dictionary pairs are used for reconstruction. In the stage of solving the sparse coefficient,the effect of noise is removed by introducing sparse coding noise,a good estimation of the sparse-coding coefficient of the original image is obtained by using the non-local self similarity of the image,and then the sparse-coding coefficients of the observed image are concentrated into these estimates. Experiments show that compared with ASDS algorithm and SSM algorithm,this algorithm has better reconstruction results,and gets richer image details and clearer edges.
引文
[1] YU J,GAO X,TAO D,et al. A unified learning framework for single image super-resolution[J]. IEEE Transactions on Neural networks and Learning systems,2014,25(4):780-792.
[2] XIE C,ZENG W,JIANG S,et al. Multiscale self-similarity and sparse representation based single image super-resolution[J]. Neurocomputing,2017,260:92-103.
[3] ZEYDE R,ELAD M,PROTTER M. On single image scale-up using sparse-representations[C]//International Conference on Curves and Surfaces. Berlin:Springer,2010:711-730.
[4] YANG J,WRIGHT J,HUANG T S,et al. Image Super-Resolution Via Sparse Representation[J]. IEEE Transactions on Image Processing,2010,19(11):2861-2873.
[5] ZHANG K,TAO D,GAO X,et al. Learning multiple linear mappings for efficient single image super-resolution[J]. IEEE Transactions on Image Processing,2015,24(3):846-861.
[6] GLASNER D,BAGON S,IRANI M. Super-resolution from a single image[C]//2009 IEEE 12th International Conference on Computer Vision. 2009:349-356.
[7]刘冬雪,秦晅,杨欣,等.一种改进的在线字典学习图像超分辨率重建算法[J].云南民族大学学报(自然科学版),2018(02):140-146.
[8] ELAD M,YAVNEH I. A plurality of sparse representations is better than the sparsest one alone[J]. IEEE Transactions on Information Theory,2009,55(10):4701-4714.
[9]何苗,王保云,盛伟,等.隐式低秩表示联合稀疏表示的人脸识别方法[J].云南师范大学学报(自然科学版),2017,37(01):43-51.
[10] ISA N A M,SALAMAH S A,NGAH U K. Adaptive fuzzy moving K-means clustering algorithm for image segmentation[J].IEEE Transactions on Consumer Electronics,2010,55(4):2145-2153.
[11] ROSYADI T,ARIF A,ACHMAD B. Classification of leukocyte images using K-Means Clustering based on geometry features[C]//2016 6th International Annual Engineering Seminar(In AES). 2016:245-249.
[12] BUADES A,COLL B,MOREL J M. A review of image denoising algorithms,with a new one[J]. Multiscale Modeling and Simulation,2005,4(2):490-530.
[13] DONG W,ZHANG L,SHI G,et al. Image deblurring and super-resolution by adaptive sparse domain selection and adaptive regularization[J]. IEEE Transactions on Image Processing,2011,20(7):1838-1857.
[14]潘宗序,禹晶,胡少兴,等.基于多尺度结构自相似性的单幅图像超分辨率算法[J].自动化学报,2014,40(4):594-603.
[2] XIE C,ZENG W,JIANG S,et al. Multiscale self-similarity and sparse representation based single image super-resolution[J]. Neurocomputing,2017,260:92-103.
[3] ZEYDE R,ELAD M,PROTTER M. On single image scale-up using sparse-representations[C]//International Conference on Curves and Surfaces. Berlin:Springer,2010:711-730.
[4] YANG J,WRIGHT J,HUANG T S,et al. Image Super-Resolution Via Sparse Representation[J]. IEEE Transactions on Image Processing,2010,19(11):2861-2873.
[5] ZHANG K,TAO D,GAO X,et al. Learning multiple linear mappings for efficient single image super-resolution[J]. IEEE Transactions on Image Processing,2015,24(3):846-861.
[6] GLASNER D,BAGON S,IRANI M. Super-resolution from a single image[C]//2009 IEEE 12th International Conference on Computer Vision. 2009:349-356.
[7]刘冬雪,秦晅,杨欣,等.一种改进的在线字典学习图像超分辨率重建算法[J].云南民族大学学报(自然科学版),2018(02):140-146.
[8] ELAD M,YAVNEH I. A plurality of sparse representations is better than the sparsest one alone[J]. IEEE Transactions on Information Theory,2009,55(10):4701-4714.
[9]何苗,王保云,盛伟,等.隐式低秩表示联合稀疏表示的人脸识别方法[J].云南师范大学学报(自然科学版),2017,37(01):43-51.
[10] ISA N A M,SALAMAH S A,NGAH U K. Adaptive fuzzy moving K-means clustering algorithm for image segmentation[J].IEEE Transactions on Consumer Electronics,2010,55(4):2145-2153.
[11] ROSYADI T,ARIF A,ACHMAD B. Classification of leukocyte images using K-Means Clustering based on geometry features[C]//2016 6th International Annual Engineering Seminar(In AES). 2016:245-249.
[12] BUADES A,COLL B,MOREL J M. A review of image denoising algorithms,with a new one[J]. Multiscale Modeling and Simulation,2005,4(2):490-530.
[13] DONG W,ZHANG L,SHI G,et al. Image deblurring and super-resolution by adaptive sparse domain selection and adaptive regularization[J]. IEEE Transactions on Image Processing,2011,20(7):1838-1857.
[14]潘宗序,禹晶,胡少兴,等.基于多尺度结构自相似性的单幅图像超分辨率算法[J].自动化学报,2014,40(4):594-603.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |