基于双向2DPCA算法的高分五号卫星图像降维研究  被引量:2

Study on Dimensionality Reduction of GF-5 Satellite Image Based on(2D)2 PCA

在线阅读下载全文

作  者:何春 郭科[2] HE Chun;GUO Ke(Education and Information Technology Center,China West Normal University,Nanchong Sichuan 637002,China;Geomathematics Key Laboratory of Sichuan Province,Chengdu University of Technology,Chengdu Sichuan 610059,China)

机构地区:[1]西华师范大学教育信息技术中心,四川南充637009 [2]数学地质四川省重点实验室(成都理工大学),四川成都610059

出  处:《乐山师范学院学报》2020年第12期1-7,共7页Journal of Leshan Normal University

基  金:数学地质四川省重点实验室开放基金(scsxdz2018yb08);西华师范大学基本科研业务费(19D043)。

摘  要:对高光谱图像降维历来是遥感数据处理研究热点之一,而将双向2DPCA算法用于高光谱图像降维的研究却见刊极少。另外,目前对高分五号卫星图像的研究也相对较少。针对此现状,文章基于双向2DPCA算法,对高分五号卫星图像进行了降维研究,该算法能够从行和列两个方向对遥感图像进行二维主成分分析,并提取有效主成分,以达到双向降维的目的。从主成分尺寸、图像压缩率、峰值信噪比以及重构图像等指标对2DPCA算法与双向2DPCA算法进行了对比实验和数据分析。实验结果表明,在以上四个图像降维指标的表现上,双向2DPCA算法都能比2DPCA算法获得更好的降维数据,由此得出结论,双向2DPCA算法对高分五号卫星图像具有良好的降维效果。Dimensionality reduction on hyperspectral images is one of the research hotspots of remote sensing data processing.However,the research on application of two-directional two-dimensional PCA algorithm on dimensionality reduction of hyperspectral images is seldom seen in the journal.In addition,there are relatively few studies on the satellite image of GF-5.In view of this situation,this paper studied the dimensionality reduction of GF-5 satellite image based on(2D)2 PCA algorithm.This algorithm conducted two-dimensional principal component analysis of the image from the row and column directions,extracted the effective principal component,and achieved the goal of two-directional dimensionality reduction.The experimental results show that the algorithm has a satisfactory effect on the dimensionality reduction of GF-5 satellite images in terms of image compression rate,peak signal-to-noise ratio and image reconstruction performance.The study in this paper also has a promotion effect on other GF-5 satellite images.

关 键 词:高光谱图像 双向2DPCA 数据降维 投影矩阵 图像重构 

分 类 号:TP751.1[自动化与计算机技术—检测技术与自动化装置]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象