基于加权Boosting的核偏最小二乘图像超分辨率重建

李小燕; 和红杰; 尹忠科; 陈帆

doi:10.3724/SP.J.1146.2011.01191

基于加权Boosting的核偏最小二乘图像超分辨率重建

doi: 10.3724/SP.J.1146.2011.01191

基金项目:

国家自然科学基金(60970122)，教育部博士点基金(20090184120021)，中央高校基本科研业务专项基金(SWJTU09CX039, SWJTU10CX09)和四川省科技创新苗子工程项目(2011-013)资助课题

计量
- 文章访问数: 2657
- HTML全文浏览量: 93
- PDF下载量: 947
- 被引次数: 0
出版历程
- 收稿日期: 2011-11-16
- 修回日期: 2012-03-26
- 刊出日期: 2012-07-19

Image Super-resolution Reconstruction Based on Kernel Partial Least Squares and Weighted Boosting

摘要

摘要: 核偏最小二乘(KPLS)算法对每个图像块选用全部主元成分进行图像重建，导致图像超分辨率算法的计算量大。兼顾图像重建质量和时间效率，该文提出一种加权Boosting的图像超分辨率重建算法。为自适应地选取每个图像块主元成分的最佳数目，利用加权Boosting原理对KPLS回归预测量进行补偿，推导给出补偿权重系数的数学表达式。讨论不同Boosting阈值情况下的重建性能，在合适的下，选取出主元成分的最佳数目m更好地满足KPLS回归模型的精度要求。实验结果表明，该文算法的超分辨率重建质量优于传统算法。
- 图像超分辨率重建 /
- 加权Boosting /
- 核偏最小二乘(KPLS) /
- Boosting阈值 /
- 主元成分
Abstract: The Kernel Partial Least Squares (KPLS) method has a large calculation since it uses all the principal components for each image block. To consider reconstruction quality and time efficiency, a weighted Boosting based algorithm is proposed in this paper. To choose adaptively the best number of principal components for each image block, the estimator in KPLS prediction model is performed for compensation. The weight coefficient expression of compensation is deduced. The reconstruction effects in different Boosting threshold are discussed. With an appropriate threshold, the chosen best number of principal components can better satisfy KPLS regression model accuracy. Experimental results demonstrate that the proposed method outperforms the conventional methods in super-resolution reconstructed quality.
- Image super-resolution reconstruction /
- Weighted Boosting /
- Kernel Partial Least Squares (KPLS) /
- Boosting threshold /
- Principal component

HTML全文

参考文献(1)

Park S C, Park M K, and Kang M G. Super-resolution image reconstruction: a technical overview[J].IEEE Signal Processing Magazine.2003, 20(3):21-36[2]杨浩, 高建坡, 吴镇扬. 一种新的图像配准和超分辨率重建算法[J].电子与信息学报.2008, 30(1):168-171浏览[3]Yang Hao, Gao Jian-po, and Wu Zhen-yang. A new algorithm for image registration and super-resolution reconstruction[J].Journal of Electronics Information Technology.2008, 30(1):168-171[5]Glasner D.[J].Bagon S, and Irani M. Super-resolution from a single image[C]. IEEE 12th International Conference on Computer Vision, Kyoto, Japan.2009,:-[6]乔建苹, 刘琚, 闫华, 等. 基于Log-WT的人脸图像超分辨率重建[J].电子与信息学报.2008, 30(6):1276-1280浏览[7]Qiao Jian-ping, Liu Ju, Yan Hua, et al.. A Log-WT based super-resolution algorithm[J].Journal of Electronics Information Technology.2008, 30(6):1276-1280[8]Chan T M, Zhang J, Pu J, et al.. Neighbor embedding based super-resolution algorithm through edge detection and feature selection[J].Pattern Recognition Letters.2009, 30(5):494-502[9]Yang J, Wright J, Huang T, et al.. Image super-resolution via sparse representation[J].IEEE Transactions on Image Processing.2010, 19(11):2861-2873[10]李民, 李世华, 李小文, 等. 非局部联合稀疏近似的超分辨率重建算法[J].电子与信息学报.2011, 33(6):1407-1412浏览[11]Li Min, Li Shi-hua, Li Xiao-wen, et al.. Super-resolution reconstruction algorithm based on non-local simultaneous sparse approximation[J].Journal of Electronics Information Technology.2011, 33(6):1407-1412[12]胡宇, 赵保军, 沈庭芝, 等. 基于偏最小二乘的人脸超分辨率重构[J]. 北京理工大学学报, 2010, 30(9): 1098-1101.[14]Wu W, Liu Z, and He X. Learning-based super resolution using kernel partial least squares[J].Image and Vision Computing.2011, 29(6):394-406[15]王惠文, 吴载斌, 孟洁. 偏最小二乘回归的线性与非线性方法[M]. 北京: 国防工业出版社, 2006: 97-104, 215-225.[17]Rosipal R and Krmer N. Overview and recent advances in partial least squares[J].Lecture Notes in Computer Science.2006, 3940:34-51[18]Chen S, Wang J, Ouyang Y, et al.. Boosting part-sense multi-feature learners toward effective object detection[J].Computer Vision and Image Understanding.2011, 115(3):364-374[19]Chang C C. A boosting approach for supervised Mahalanobis distance metric learning[J].Pattern Recognition.2012, 45(2):844-862[20]Suresh S, Sundararajan N, and Saratchandran P. Risk- sensitive loss functions for sparse multi-category classification problems[J].Information Sciences.2008, 178(15):2621-2638[21]Gao X, Zhang K, Tao D, et al.. Joint learning for single image super-resolution via coupled constraint[J].IEEE Transactions on Image Processing.2012, 21(2):469-480[22]Wang Z, Bovik A C, Sheikh H R, et al.. Image quality assessment: from error measurement to structural similarity[J].IEEE Transactions on Image Processing.2004, 13(4):600-612

施引文献

资源附件(0)

访问统计