[1]齐永锋,陈静,火元莲,等.基于多尺度卷积神经网络的高光谱图像分类算法[J].红外技术,2020,42(9):855-862.[doi:10.11846/j.issn.1001_8891.202009007]
 QI Yongfeng,CHEN Jing,HUO Yuanlian,et al.Hyperspectral Image Classification Algorithm Based on Multiscale Convolutional Neural Network[J].Infrared Technology,2020,42(9):855-862.[doi:10.11846/j.issn.1001_8891.202009007]
点击复制

基于多尺度卷积神经网络的高光谱图像分类算法
分享到:

《红外技术》[ISSN:1001-8891/CN:CN 53-1053/TN]

卷:
42卷
期数:
2020年第9期
页码:
855-862
栏目:
出版日期:
2020-09-23

文章信息/Info

Title:
Hyperspectral Image Classification Algorithm Based on Multiscale Convolutional Neural Network
文章编号:
1001-8891(2020)09-0855-08
作者:
齐永锋1陈静1火元莲2李发勇1
1. 西北师范大学 计算机科学与工程学院;
2. 西北师范大学 物理与电子工程学院
Author(s):
1. College of Computer Science and Engineering, Northwest Normal University;
2. College of Physics and Electronic Engineering, Northwest Normal University

关键词:
高光谱图像等距特征映射多尺度卷积神经网络分类
Keywords:
hyperspectral image isometric feature mapping multiscale convolutional neural network classification
分类号:
TP391.9
DOI:
10.11846/j.issn.1001_8891.202009007
文献标志码:
A
摘要:
为了提高高光谱图像的分类精度,提出了一种基于多尺度卷积神经网络的高光谱图像分类算法。首先,利用等距特征映射算法处理高光谱数据,以挖掘数据的非线性特性,保持数据点的内在几何性质;然后,构建以标记像元为中心的训练图像块,训练多尺度卷积神经网络;最后,利用softmax分类器预测测试像元的标签。提出的方法在Indian Pines、University of Pavia和Salinas scene高光谱遥感数据集上进行分类实验,并与CNN、R-PCA CNN、CNN-PPF、CD-CNN等算法进行性能比较。实验结果表明,在3个数据集上提出的方法的总体识别精度分别达到98.51%、98.64%和99.39%,与CNN算法相比分别提高了约8.35%、6.37%和7.81%。本文提出的方法无论是在分类精度还是Kappa系数上都优于另外4种方法,是一种较好的高光谱遥感数据分类方法。
Abstract:
To improve the classification accuracy of hyperspectral remote sensing images, a classification algorithm based on a multiscale convolutional neural network (CNN) is proposed. First, an isometric feature mapping algorithm was used to process hyperspectral data, to mine the nonlinear characteristics of the data and maintain the intrinsic geometric properties of data points. Second, training image blocks centered on labeled pixels were constructed, after which the multiscale CNNs were trained. Finally, the Softmax classifier was used to predict the label of the test pixel. The proposed method performed classification experiments on the Indian Pines, University of Pavia, and Salinas scene hyperspectral remote sensing datasets, and its performance was compared with a CNN, randomized principal component analysis (R-PCA CNN), a deep CNN with pixel-pair features (CNN-PPF), a cross-domain CNN (CD-CNN), and other algorithms. The experimental results showed that the overall recognition accuracy of the proposed method for the three datasets was 98.51%, 98.64%, and 99.39%, respectively, which was 8.35%, 6.37%, and 7.81% higher than that of the CNN algorithm, respectively. The proposed method performed better than the other four methods studied, in terms of both classification accuracy and Kappa coefficient, providing a superior method for hyperspectral remote sensing data classification.

参考文献/References:

[1]? LI Shutao, SONG Weiwei, FANG Leyuan, et al. Deep Learning for Hyperspectral Image Classification: An Overview[J]. IEEE Transactions on Geoscience and Remote Sensing, 2019, 57(9): 6690-6709.
[2]? TAN Kun, MA Weibo, WU Fuyu, et al. Random forest-based estimation of heavy metal concentration in agricultural soils with hyperspectral sensor data[J]. Environmental Monitoring & Assessment, 2019, 191(7): 1-14.?
[3]? CAO Xianghai, WANG Xiaozhen, WANG Da, et al. Densely connected deep random forest for hyperspectral imagery classification[J]. International Journal of Remote Sensing, 2018, 40(3): 1-16.
[4]? ZHANG Cuifen, YANG Xiaoxia, HAO Lina, et al. Lithological classification by synergizing hyperspectral Hyperion and high resolution WorldView-2 satellite images[J]. Journal of Chengdu University of Technology: Science & Technology Edition, 2017, 44(5): 613-622.
[5]? 杜培军, 夏俊士, 薛朝辉, 等. 高光谱遥感影像分类研究进展[J]. 遥感学报, 2016, 20(2): 236-256.
DU Peijun, XIA Junshi, XUE Chaohui, et al. Research progress on classification of hyperspectral remote sensing images[J]. Journal of Remote Sensing, 2016, 20(2): 236-256.
[6]? 刘靖, 赵逢禹. 高维数据降维技术及研究进展[J]. 电子科技, 2018, 31(3): 36-38.
LIU Jing, ZHAO Fengqi. High-dimensional data dimension reduction technology and research progress[J]. Electronic Technology, 2018, 31(3): 36-38.
[7]? LIU Q, CAI Y, JIANG H, et al. Traffic state prediction using ISOMAP manifold learning[J]. Physica A: Statistical Mechanics and its Applications, 2018, 506(15): 532-541.
[8]? 王程锦, 王秀友, 林玉娥. 基于内类子空间学习的局部线性嵌入算法[J]. 阜阳师范学院学报: 自然科学版, 2019, 36(2): 52-56.
WANG Chengjin, WANG Xiuyou, LIN Yuxi. Local Linear Embedding Algorithm Based on Inner Class Subspace Learning[J]. Journal of Fuyang Teachers College: Natural science, 2019, 36(2): 52-56.
[9]? LI S, WANG Z, LI Y. Using Laplacian Eigenmap as Heuristic Information to Solve Nonlinear Constraints Defined on a Graph and Its Application in Distributed Range-Free Localization of Wireless Sensor Networks[J]. Neural Processing Letters, 2013, 37(3): 411-424.
[10]? A Krizhevsky, I Sutskever, G E Hinton. Image Net classification with deep convolutional neural networks[C]//Proc. Adv. Neural Inf. Process. Syst. (NIPS), 2012, 25(2): 1097-1105.
[11]? 王博. 基于深度学习的低信息量图片物体检测算法研究[J]. 价值工程, 2019, 38(24): 256-259.
WANG Bo. Research on Low-information Image Object Detection Algorithm Based on Deep Learning[J]. Value Engineering, 2019, 38(24): 256-259.
[12]? 崔莹. 深度学习在文本表示及分类中的应用研究[J]. 电脑知识与技术, 2019, 15(16): 174-177.
CUI Ying. Application of Deep Learning in Text Representation and Classification[J]. Computer Knowledge and Technology, 2019, 15(16): 174-177.
[13]? ZHANG L, ZHANG L, DU B. Deep learning for remote sensing data: A technical tutorial on the state of the art[J]. IEEE Geosci. Remote Sens. Mag., 2016, 4(2): 22-40.?
[14]? CHENG G, ZHOU P, HAN J. Learning rotation-invariant convolutional neural networks for object detection in VHR optical remote sensing images[J]. IEEE Trans. Geosci. Remote Sens., 2016, 54(12): 7405-7415.?
[15]? YAO X, HAN J, CHENG G, et al. Semantic annotation of high-resolution satellite images via weakly supervised learning[J]. IEEE Trans. Geosci. Remote Sens., 2015, 4(6): 3660-3671.?
[16]? CHEN Y, LIN Z, ZHAO X, et al. Deep learning-based classification of hyperspectral data[J]. IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens., 2014, 7(6): 2094-2107.?
[17]? LIU Y, CAO G, SUN Q, et al. Hyperspectral classification via deep networks and superpixel segmentation[J]. Int. J. Remote Sens., 2015, 36(13): 3459-3482.?
[18]? MA X, WANG H, GENG J. Spectral-spatial classification of hyperspectral image based on deep auto-encoder[J]. IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens., 2016, 9(9): 4073-4085.?
[19]? CHEN Y, ZHAO X, JIA X. Spectral-spatial classification of hyperspectral data based on deep belief network[J]. IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens., 2015, 8(6): 2381-2392.
[20]? ZHONG P, GONG Z, LI S, et al. Learning to diversify deep belief networks for hyperspectral image classification[J]. IEEE Trans. Geosci. Remote Sens., 2017, 55(6): 3516-3530.?
[21]? 杨建功, 江西莉, 刘侍刚. 融合谱-空域信息的DBN高光谱图像分类方法[J]. 西安电子科技大学学报, 2019, 46(3): 109-115.
YANG Jiangong, LI Jiangxi, LIU Shigang. DBN hyperspectral image classification method based on fusion spectrum-airspace information[J]. Journal of Xidian University, 2019, 46(3): 109-115.
[22]? CHEN Y, JIANG H, LI C, et al. Deep feature extraction and classification of hyperspectral images based on convolutional neural networks[J]. IEEE Trans. Geosci. Remote Sens., 2016, 54(10): 6232-6251.?
[23]? LI W, WU G, ZHANG F, et al. Hyperspectral image classification using deep pixel-pair features[J]. IEEE Trans. Geosci. Remote Sens., 2017, 55(2): 844-853.?
[24]? 丁玲, 唐娉, 李宏益. 基于ISOMAP的高光谱遥感数据的降维与分类[J]. 红外与激光工程, 2013, 42(10): 2708-2711.
DING Ling, TANG Wei, LI Hongyi. Dimensionality and classification of hyperspectral remote sensing data based on ISOMAP[J]. Infrared and Laser Engineering, 2013, 42(10): 2708-2711.
[25]? Enayattabar M, Ebrahimnejad A, Motameni H. Dijkstra algorithm for shortest path problem under interval-valued Pythagorean fuzzy environment[J]. Complex & Intelligent Systems, 2019, 5(2): 93-100.?
[26]? 唐文俊. 基于流行学习的数据降维的研究[D]. 广东: 广东工业大学, 2012.
TANG Wenjun. Research on data dimensionality reduction based on popular learning[D]. Guangdong: Guangdong University of Technology, 2012.
[27]? 吴从中, 陈曦, 季栋, 等. 结合深度残差学习和感知损失的图像去噪[J]. 中国图象图形学报, 2018, 23(10): 1483-1491.
WU Congzhong, CHEN Wei, JI Dong, et al. Image Denoising Combined with Deep Residual Learning and Perceptual Loss[J]. Journal of Image and Graphics, 2018, 23(10): 1483-1491.
[28]? HU W, HUANG Y, LI W, et al. Deep convolutional neural networks for hyperspectral image classification[J]. Journal of Sensors, 2015, 2015(2): 1-12.
[29]? K Makantasis, K Karantzalos, A Doulamis, et al. Deep supervised learning for hyperspectral data classification through convolutional neural networks[C]//IEEE International Geoscience and Remote Sensing Symposium, Milan, Italy, 2015: 4959-4962.
[30]? LI W, WU G, ZHANG F, et al. Hyperspectral image classification using deep pixel-pair features[J]. IEEE Trans. Geosci. Remote Sens., 2017, 55(2): 844-853.
[31]? LEE H, KWON H. Going deeper with contextual CNN for hyperspectral image classification[J]. IEEE Trans. Image Process., 2017, 26(10): 4843-4855.
[32]? ZHONG M, LI W, DU Q. Diverse region-based CNN for hyperspectral image classification[J]. IEEE Transactions on Image Processing A Publication of the IEEE Signal Processing Society, 2018, 27(6): 2623-2634.

相似文献/References:

[1]张翔.基于光谱特征增强的高光谱图像地物目标识别[J].红外技术,2010,32(12):717.
 Target Recognition of Hyperspectral Image Based?on Enhanced Spectral Characteristics[J].Infrared Technology,2010,32(9):717.
[2]何高攀,杨桄,张筱晗,等.基于端元提取的高光谱异常目标检测[J].红外技术,2015,37(十):836.[doi:10.11846/j.issn.1001_8891.201510006]
 HE Gao-pan,YANG Guang,ZHANG Xiao-han,et al.Anomaly Detection Algorithm Based on Endmember Extraction in Hyperspectral Imagery[J].Infrared Technology,2015,37(9):836.[doi:10.11846/j.issn.1001_8891.201510006]
[3]陈海挺.基于非下采样Contourlet变换的异常检测SVDD算法[J].红外技术,2016,38(1):047.[doi:10.11846/j.issn.1001_8891.201601008]
 CHEN Haiting.Anomaly Detection SVDD Algorithm Based on Nonsubsampled Contourlet Transform[J].Infrared Technology,2016,38(9):047.[doi:10.11846/j.issn.1001_8891.201601008]
[4]向英杰,张俭峰,杨 桄,等.基于混合噪声估计的高光谱图像异常检测方法[J].红外技术,2017,39(8):734.[doi:10.11846/j.issn.1001_8891.201708011]
 XIANG Yingjie,ZHANG Jianfeng,YANG Guang,et al.A Mixed-Noise Estimation-Based Anomaly Detection Method for Hyperspectral Image [J].Infrared Technology,2017,39(9):734.[doi:10.11846/j.issn.1001_8891.201708011]
[5]向英杰,杨 桄,张俭峰,等. 采用分段行-列核2DPCA的高光谱图像降维[J].红外技术,2017,39(12):1107.[doi:10.11846/j.issn.1001_8891.201712007]
 XIANG Yingjie,YANG Guang,ZHANG Jianfeng,et al. Dimensionality Reduction for Hyperspectral Image Using a Segmented Row-column Kernel Two-dimensional Principal Component Analysis Method[J].Infrared Technology,2017,39(9):1107.[doi:10.11846/j.issn.1001_8891.201712007]
[6]徐杭威,赵 壮,岳 江,等. 一种基于归一化光谱向量的高光谱图像实时性非监督分类方法[J].红外技术,2018,40(4):362.[doi:10.11846/j.issn.1001_8891.201804010]
 XU Hangwei,ZHAO Zhuang,YUE Jiang,et al.Real-time Unsupervised Classification Method of Hyperspectral Images Based on the Normalized Spectral Vector [J].Infrared Technology,2018,40(9):362.[doi:10.11846/j.issn.1001_8891.201804010]
[7]徐 君,王旭红,杨 勇,等.一种基于光谱角空间变换的高光谱图像分割方法[J].红外技术,2018,40(10):1013.[doi:10.11846/j.issn.1001_8891.201810014]
 XU Jun,WANG Xuhong,YANG Yong,et al.A Hyperspectral Image Segmentation Method Based on Spectral Angular Space Transformation [J].Infrared Technology,2018,40(9):1013.[doi:10.11846/j.issn.1001_8891.201810014]
[8]严 阳,华文深,张 炎,等.可见-近红外高光谱伪装目标特性分析[J].红外技术,2019,41(2):171.[doi:10.11846/j.issn.1001_8891.2019020011]
 YAN Yang,HUA Wenshen,ZHANG Yan,et al.Visible Near-infrared Hyperspectral Camouflage Target Characteristic Analysis[J].Infrared Technology,2019,41(9):171.[doi:10.11846/j.issn.1001_8891.2019020011]
[9]张国东,周浩,方淇,等.基于栈式自编码神经网络对高光谱遥感图像分类研究[J].红外技术,2019,41(5):450.[doi:10.11846/j.issn.1001_8891.201905010]
 ZHANG Guodong,ZHOU Hao,FANG Qi,et al.Classification of Hyperspectral Remote Sensing Images Based on Stack Self-encoding Neural Network[J].Infrared Technology,2019,41(9):450.[doi:10.11846/j.issn.1001_8891.201905010]
[10]陈 欣,粘永健,王忠良.基于线性混合模型的高光谱图像分布式压缩感知[J].红外技术,2019,41(8):758.[doi:10.11846/j.issn.1001_8891.2019080011]
 CHEN Xin,NIAN Yongjian,WANG Zhongliang.Distributed Compressive Sensing for Hyperspectral Imaging Based on Linear Mixing Model [J].Infrared Technology,2019,41(9):758.[doi:10.11846/j.issn.1001_8891.2019080011]

备注/Memo

备注/Memo:
收稿日期:2019-11-28;修订日期:2020-09-03.
作者简介:齐永锋(1971-),男,教授,博士,主要研究方向:模式识别与数字图像,E-mail: qiyf@nwnu.edu.cn。?
基金项目:甘肃省高等学校科研项目(2016A-004);甘肃省科技计划项目(18JR3RA097)。?

更新日期/Last Update: 2020-09-18