首页 > 论文 > 红外与激光工程 > 46卷 > 12期(pp:1228001--1)

空谱结构保持的高光谱图像分类

Hyperspectral image classification based on spatial-spectral structure preserving

  • 摘要
  • 论文信息
  • 参考文献
  • 被引情况
  • PDF全文
分享:

摘要

高光谱遥感图像具有特征(波段)数多、冗余度高等特点, 因此特征选择成为高光谱分类的研究热点。针对此问题, 提出了空间结构与光谱结构同时保持的高光谱数据分类算法。考虑高光谱图像的物理特性, 首先对图像进行加权空谱重构, 使图像的空间结构信息自动融入光谱特征, 形成空谱特征集; 对利用最小二乘回归模型保存数据集的全局相似性结构的基础上, 加入局部流形结构正则项, 使挑选的特征子集更好地保存数据集的内在本质结构; 讨论了窗口大小和正则参数对分类精度的影响。对Indian Pines、PaviaU和Salinas数据集的实验表明, 该算法得到的特征子集的总体分类精度达到93.22%、96.01%和95.90%。该算法不仅充分利用了高光谱图像的空间结构信息, 而且深入挖掘了数据集的内在本质结构, 从而得到更有鉴别性的特征子集, 相比传统方法明显提高了分类精度。

Abstract

Hyperspectral remote sensing image contains the properties of much features(bands) and high redundancy, and the research of hyperspectral image classification focuses on feature selection. To overcome this problem, a hyperspectral image classification algorithm based on spatial and spectral structure preserving was proposed. Considering the physical characteristics of hyperspectral image, the weighted spatial and spectral reconstruction of the image was conducted firstly, in order to incorporate spatial structure information into the spectral feature set automatically, resulting in the spatial-spectral feature set. On the basis that the least square regression model uncovered the global similarity structure and the regularization term revealed the local manifold structure, the intrinsic structure of the spatial-spectral feature set was well preserved by the selected feature subset. The influence of window size and regularization parameter was also analyzed. The experiments on Indian Pines, PaviaU and Salinas datasets show that the classification accuracy of the proposed algorithm reaches 93.22%, 96.01% and 95.90% respectively. The proposed method not only makes full use of the spatial structure information of the hyperspectral image but also uncovers the intrinsic structure of the dataset, which contribute to select more discriminant feature subset and obtain higher classification accuracy compared with conventional methods.structure preserving

Newport宣传-MKS新实验室计划
补充资料

中图分类号:TP751

DOI:10.3788/irla201746.1228001

所属栏目:景象信息处理

基金项目:国家自然科学基金 (61401471); 中国博士后科学基金(2014M562636)

收稿日期:2017-04-07

修改稿日期:2017-05-12

网络出版日期:--

作者单位    点击查看

侯榜焕:火箭军工程大学 信息工程系, 陕西 西安 710025
姚敏立:火箭军工程大学 信息工程系, 陕西 西安 710025
贾维敏:火箭军工程大学 信息工程系, 陕西 西安 710025
沈晓卫:火箭军工程大学 信息工程系, 陕西 西安 710025
金 伟:火箭军工程大学 信息工程系, 陕西 西安 710025

联系人作者:侯榜焕(chinayouth001@aliyun.com)

备注:侯榜焕(1985-), 男, 博士生, 主要从事高光谱图像处理方面的研究。

【1】Deng Chengzhi, Zhang Shaoquan, Wang Shengqian, et al. Hyperspectral unmixing algorithm based on L1 regularization [J]. Infrared and Laser Engineering, 2015, 44(3): 1092-1097. (in Chinese)
邓承志, 张绍泉, 汪胜前, 等. L1稀疏正则化的高光谱混合像元分解算法比较[J].红外与激光工程, 2015, 44(3): 1092-1097.

【2】Zhao Huijie, Li Mingkang, Li Na, et al. A band selection method based on improved subspace partition[J]. Infrared and Laser Engineering, 2015, 44(10): 3155-3160. (in Chinese)
赵慧洁, 李明康, 李娜, 等 一种基于改进子空间划分的波段选择方法[J]. 红外与激光工程, 2015, 44(10): 3155-3160.

【3】Fang Min, Wang Jun, Wang Hongyan, et al. Feature extraction of hyperspectral remote sensing data using supervised neighbor reconstruction analysis [J]. Infrared and Laser Engineering, 2016, 45(10): 1028003. (in Chinese)
方敏, 王君, 王红艳, 等. 应用监督近邻重构分析的高光谱遥感数据特征提取[J].红外与激光工程, 2016, 45(10): 1028003.

【4】Feng Shuyi, Zhang Ning, Shen Ji, et al. Method of cloud detection with hyperspectral remote sensing image based on the reflective characteristic [J]. Chinese Optics, 2015, 8(2): 199-205. (in Chinese)
冯书谊, 张宁, 沈霁, 等. 基于反射率特性的高光谱遥感图像云检测方法研究[J]. 中国光学, 2015, 8(2): 199-205.

【5】Liu X W, Wang L, Zhang J, et al. Global and local structure preservation for feature selection [J]. IEEE Transactions on Neural Network and Learning System, 2014, 25(6): 1083-1095.

【6】Nie F P, Huang H, Cai X, et al. Efficient and robust feature selection via joint -norms minimization [C]//Proceedings of Advances in Nerual Information Processing System, 2010: 1813-1821.

【7】Zhu X F, Li X L, Zhang S C, et al. Robust joint graph sparse coding for unsupervised spectral feature selection [J]. IEEE Transactions on Neural Network and Learning System, 2017, 28(6): 1263-1275.

【8】Zhang Q, Tian Y, Yang Y P, et al. Automatic spatial-spectral feature selection for hyperspectral image via discriminative sparse multimodal learning [J]. IEEE Transactions on Geoscience and Remote Sensing, 2015, 53(1): 261-279.

【9】Li H C, Xiang S M, Zhong Z S, et al. Multicluster spatial-spectral unsupervised feature selection for hyperspectral image classification [J]. IEEE Geoscience and Remote Sensing Letters, 2015, 12(8): 1660-1664.

【10】Zhou Y C, Peng J T, Chen C L P. Dimension reduction using spatial and spectral regularized local discriminant embedding for hyperspectral image classification[J]. IEEE Transactions on Geoscience and Remote Sensing, 2015, 53(2): 1082-1095.

【11】Huang Hong, Zheng Xinlei. Hyperspectral image classification with combination of weighted spatial-spectral and KNN [J]. Optics and Precision Engineering, 2016, 24(4): 873-880. (in Chinese)
黄鸿, 郑新磊. 加权空-谱与最近邻分类器相结合的高光谱图像分类[J]. 光学 精密工程, 2016, 24(4): 873-880.

【12】He Fang, Wang Rong, Yu Qiang, et al. Feature extraction of hyperspectral images of weighted spatial and spectral locality preserving Projection [J]. Optics and Precision Engineering, 2017, 25(1): 263-273. (in Chinese)
何芳, 王榕, 于强, 等. 加权空谱局部保持投影的高光谱图像特征提取[J]. 光学 精密工程, 2017, 25(1): 263-273.

【13】Hao Zhicheng, Wu Chuan, Yang Hang, et al. Image detail enhancement method based on multi-scale bilateral texture filter [J]. Chinese Optics, 2016, 9(4): 423-431. (in Chinese)
郝志成,吴川,杨航,等.基于双边纹理滤波的图像细节增强方法[J].中国光学, 2016, 9(4): 423-431.

【14】He X, Niyogi P. Locality preserving projections[C]//Preceedings of the 17th Annual Conference on Neural Information Processing Systems, 2004, 16: 153-160.

引用该论文

Hou Banghuan,Yao Minli,Jia Weimin,Shen Xiaowei,Jin wei. Hyperspectral image classification based on spatial-spectral structure preserving[J]. Infrared and Laser Engineering, 2017, 46(12): 1228001

侯榜焕,姚敏立,贾维敏,沈晓卫,金 伟. 空谱结构保持的高光谱图像分类[J]. 红外与激光工程, 2017, 46(12): 1228001

被引情况

【1】闫敬文,陈宏达,刘 蕾. 高光谱图像分类的研究进展. 光学 精密工程, 2019, 27(3): 680-693

您的浏览器不支持PDF插件,请使用最新的(Chrome/Fire Fox等)浏览器.或者您还可以点击此处下载该论文PDF