非参数贝叶斯字典学习的遥感影像超分辨率重建

doi:10.13474/j.cnki.11-2246.2018.0199

测绘通报 ›› 2018, Vol. 0 ›› Issue (7): 5-8,12.doi: 10.13474/j.cnki.11-2246.2018.0199

非参数贝叶斯字典学习的遥感影像超分辨率重建

李丽¹, 隋立春^1,2, 康军梅¹, 王雪¹

1. 长安大学地质工程与测绘学院, 陕西西安 710054;
2. 地理国情监测国家测绘地理信息局工程中心, 陕西西安 710054

收稿日期:2017-10-23 出版日期:2018-07-25 发布日期:2018-08-02
作者简介:李丽(1987-),女,博士生,主要从事遥感影像超分辨率重建算法方面的研究。E-mail:15829779607@163.com
基金资助:
国家自然科学基金（41372330；41571346）；国家自然科学基金青年科学基金（41601345）

Remote Sensing Images Super-resolution Reconstruction Based on Bayesian Nonparametric Dictionary Learning

LI Li¹, SUI Lichun^1,2, KANG Junmei¹, WANG Xue¹

1. College of Geology Engineering and Geomatics, Chang'an University, Xi'an 710054, China;
2. National Geographic Condition Monitoring National Mapping Geographic Information Bureau Engineering Center, Xi'an 710054, China

Received:2017-10-23 Online:2018-07-25 Published:2018-08-02

摘要/Abstract

摘要： 为充分利用样本及参数的先验信息，对Yang提出的基于稀疏表示的超分辨率重建算法进行了改进，提出了一种基于非参数贝叶斯字典学习的单幅遥感影像超分辨率重建方法。该方法利用Beta-Bernoulli过程进行字典学习，建立字典元素和各参数的概率分布模型，并通过Gibbs进行迭代抽样构成马尔科夫链，用其平稳分布来近似字典元素及各参数的后验分布，最后由低分辨率影像及高分辨率字典的后验分布重建出高分辨率遥感影像。对比双线性、双三次插值及Yang的方法，该算法在平均峰值信噪比方面分别提高了3.29、1.79、0.17 dB，在平均ERGAS方面分别降低了0.78、0.37、0.02 dB。该算法因加入了更多的先验信息，可以使重建影像提供更多高频细节信息，具有一定的普适性。

关键词: 超分辨率重建, 遥感影像, 稀疏表示, 非参数贝叶斯, 概率分布

Abstract: Aiming at taking advantage of the prior information of the samples and the parameters,a single remote sensing image super-resolution method based on nonparametric Bayesian dictionary learning is proposed,modified on the sparse representation algorithm presented by Yang.The Beta-Bernoulli process was used to learn the dictionary,establishing the probability distribution models of dictionary elements and parameters,and the stationary distribution of the Markov chain obtained by Gibbs iteration sampling was used to approximate their posterior.Finally,the related high resolution images were reconstructed by the posterior and the low resolution images.Compared with the bilinear,the bicubic and the super-resolution algorithm based on sparse representation, the average peak signal-to-noise ratio obtained using the method presented in this paper were increased by 3.29,1.79, 0.17 dB,respectively.the average ERGAS were decreased by 0.78, 0.37, 0.02 dB,respectively.Because of more prior added,the reconstruction image produced using the proposed method could provide more high frequency details,performing universally.

Key words: super-resolution, remote sensing images, sparse representation, nonparametric Bayesian, probability distribution

中图分类号:

P237

李丽, 隋立春, 康军梅, 王雪. 非参数贝叶斯字典学习的遥感影像超分辨率重建[J]. 测绘通报, 2018, 0(7): 5-8,12.

LI Li, SUI Lichun, KANG Junmei, WANG Xue. Remote Sensing Images Super-resolution Reconstruction Based on Bayesian Nonparametric Dictionary Learning[J]. 测绘通报, 2018, 0(7): 5-8,12.

参考文献

[1] 刘帅,朱亚杰,薛磊.一种结合稀疏表示和纹理分块的遥感影像超分辨率方法[J].武汉大学学报(信息科学版),2015,40(5):578-582.
[2] 徐青.遥感影像融合与分辨率增强技术[M].北京:科学出版社,2007:1-34.
[3] HAMIS J.Diffraction and Resolving Power[J].Journal of the Optical Society of America(1917-1983),1964,54(7):931-933.
[4] 钟九生.基于稀疏表示的光学遥感影像超分辨率重建算法研究[D].南京:南京师范大学,2013:1-17.
[5] 沈焕锋,李平湘,张良培,等.图像超分辨率重建技术与方法综述[J].光学技术,2009,35(2):194-199.
[6] POLATKAN G,ZHOU M Y,CARIN L,et al.A Bayesian Nonparametric Approach to Image Super-resolution[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2015,37(2):346-358.
[7] FREEMAN W T,JONES T R,PASZTOR E C.Example-based Super-resolution[J].IEEE Computer Graphics and Applications,2013,22(2):56-65.
[8] CHANG H,YEUNG D Y,XIONG Y.Super-resolution through Neighbor Embedding[C]//Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.Washington D C:IEEE,2004:275-282.
[9] YANG J,WRIGHT J,HUANG T,et al.Image Super-resolution as Sparse Representation of Raw Image Patches[C]//IEEE Conference on Computer Vision and Pattern Recognition.Anchorage:IEEE,2008:1-8.
[10] YANG J,WRIGHT J,HUANG T,et al.Image Super-resolu-tion via Sparse Representation[J].IEEE Transactions on Image Processing,2010,19(11):2861-2873.
[11] ZHOU M,CHEN H,PAISLEY J,et al.Non-parametric Bayesian Dictionary Learning for Sparse Image Representa-tions[J].International Conference on Neural Information Processing Systems,2295-2303.
[12] RODRIGUEZ A,DUNSON D B.Nonparametric Bayesian Models through Probit Stick-breaking Processes[J].Bayesian Analysis,2011,6(1):145-178.
[13] KNOWLES D,GHAHRAMANI Z.Infinite Sparse Factor Analysis and Infinite Independent Components Analysis[C]//International Conference on Independent Component Analysis and Signal Separation.[S.l.]:Springer-Verlag,2007:381-388.
[14] PAISLEY J,CARIN L.Nonparametric Factor Analysis with Beta Process Priors[C]//International Conference on Machine Learning.New York:ACM,2009:777-784.
[15] ZHOU M,YANG H,SAPIRO G,et al.Dependent Hierarchical Beta Process for Image Interpolation and Denoising[C]//Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics.[S.l.]:Aistats,2011:883-891.
[16] ZHOU M,CHEN H,PAISLEY J,et al.Nonparametric Bayesian Dictionary Learning for Analysis of Noisy and Incomplete Images[J].IEEE Transactions on Image Processing a Publication of the IEEE Signal Processing Society,2012,21(1):130-144.

非参数贝叶斯字典学习的遥感影像超分辨率重建

Remote Sensing Images Super-resolution Reconstruction Based on Bayesian Nonparametric Dictionary Learning

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价

[1]	郑艳, 何欢, 卜丽静, 金鑫. 自相似性和边缘保持分解的超分辨率重建算法[J]. 测绘通报, 2022, 0(7): 54-59.
[2]	李茂林, 闫庆武, 仲晓雅. 一种复合型校正植被信息的方法[J]. 测绘通报, 2022, 0(6): 49-54142.
[3]	张龚泉, 岳建平, 刘胜男, 王海龙, 杨智翔, 周航宇. 基于遥感影像的鄱阳湖地形提取方法[J]. 测绘通报, 2022, 0(5): 62-66,73.
[4]	潘建平, 付占宝, 程卫华, 陆世安, 李鑫, 孙博文, 黄桂萍. 基于地表类型的山地海绵城市生态建设评估[J]. 测绘通报, 2022, 0(5): 84-88.
[5]	陈智朗, 付振华, 朱紫阳, 王慧慧, 刘沁雯, 杨钰灵, 许耿然. 基于HRNet的高分辨率遥感影像建筑物变化信息提取[J]. 测绘通报, 2022, 0(5): 126-132.
[6]	刘小玉, 刘扬, 杜明义, 张敏, 贾竞珏, 杨恒. 基于DeeplabV3+的建筑垃圾堆放点识别[J]. 测绘通报, 2022, 0(4): 16-19,43.
[7]	王雪, 梁珂, 隋立春, 钟棉卿, 朱剑锋. 膨胀卷积与金字塔表达的深度学习模型用于农村建筑物提取[J]. 测绘通报, 2022, 0(4): 61-65.
[8]	高鸣, 周鑫鑫, 刘琦, 杨光迪, 吴长彬. 深度学习网络支持下的农房侵占耕地自动化监测[J]. 测绘通报, 2022, 0(3): 47-53.
[9]	王胜利, 朱寿红, 蒋毅. 基于CNN模型迁移的OLI影像光伏电池板场景识别[J]. 测绘通报, 2022, 0(2): 5-9.
[10]	杨素妨, 杜林. 多特征流形鉴别嵌入的高分辨率遥感影像分类[J]. 测绘通报, 2021, 0(9): 37-42.
[11]	耿万轩, 周维勋, 金双根. 基于LDCNN特征提取的多核SVM高分辨率遥感影像场景分类[J]. 测绘通报, 2021, 0(8): 14-21.
[12]	苏艺凡, 党建武, 王阳萍, 杨景玉. 改进的区间二型模糊聚类遥感影像变化检测[J]. 测绘通报, 2021, 0(7): 44-51,58.
[13]	聂倩, 七珂珂, 赵艳福. 融入超像素分割的高分辨率影像面向对象分类[J]. 测绘通报, 2021, 0(6): 44-49.
[14]	侯翘楚, 李必军, 蔡毅. 高分辨率遥感影像的车道级高精地图要素提取[J]. 测绘通报, 2021, 0(3): 38-43.
[15]	张伟良, 刘琦, 吴长彬, 杨光迪. 基于孪生神经网络的土地利用现状年度变化检测[J]. 测绘通报, 2021, 0(3): 91-95,104.