[1]王浩,张晶晶,李园园,等.基于3D卷积联合注意力机制的高光谱图像分类[J].红外技术,2020,42(3):264-271.[doi:10.11846/j.issn.1001_8891.202003009]
 WANG Hao,ZHANG Jingjing,LI Yuanyuan,et al.Hyperspectral Image Classification Based on 3D Convolution Joint Attention Mechanism[J].Infrared Technology,2020,42(3):264-271.[doi:10.11846/j.issn.1001_8891.202003009]
点击复制

基于3D卷积联合注意力机制的高光谱图像分类
分享到:

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

卷:
42卷
期数:
2020年第3期
页码:
264-271
栏目:
出版日期:
2020-03-23

文章信息/Info

Title:
Hyperspectral Image Classification Based on 3D Convolution Joint Attention Mechanism

文章编号:
1001-8891(2020)05-0264-08
作者:
王浩1张晶晶1李园园1王峰2寻丽娜1
1. 安徽大学 电气工程与自动化学院 计算智能与信号处理教育部重点实验室;
2. 偏振光成像探测技术安徽省重点实验室

Author(s):
WANG Hao1ZHANG Jingjing1LI Yuanyuan1WANG Feng2XUN Lina1
1. Key Laboratory of Computational Intelligence and Signal Processing, Ministry of Education, School of Electrical Engineering and Automation, Anhui University;
2. Key Laboratory of Polarization Imaging Detection Technology

关键词:
高光谱图像分类注意力机制深度学习像素配对
Keywords:
hyperspectral image classification attention mechanism deep learning pixel-pairs
分类号:
O235
DOI:
10.11846/j.issn.1001_8891.202003009
文献标志码:
A
摘要:
由于高光谱图像存在较高的数据维数,会给分类过程带来一些困难。为了提高分类的准确率,提出了一种使用3D卷积联合注意力机制的高光谱图像分类方法。首先,将中心像素与周围相邻的其它像素进行配对,可以通过配对构成多组新的像素对,充分利用了像素之间的邻域相关性。接着,将像素对放入3D卷积联合注意力机制网络框架中进行分类,它能够对高光谱图像中的特征进行选择性的学习。最后,通过投票策略获得像素标签。实验是在两个真实的高光谱图像数据集上进行。结果表明,所提出的方法充分挖掘了高光谱图像的光谱空间特征,能有效地提高分类精度。
Abstract:
?The high data dimension of hyperspectral images causes difficulties in the classification process. To improve the accuracy of classification, a hyperspectral image classification method using a 3D convolution joint attention mechanism is proposed. First, by pairing the center pixel with other pixels adjacent to it, it can form multiple sets of new pixel-pairs, and the neighborhood correlation between the pixels can be fully utilized. Then, the pixel-pairs are classified into the 3D convolution joint attention mechanism network framework, which can selectively learn the features in the hyperspectral image. Finally, the pixel label is obtained through the voting strategy. An experiment was carried out on two real hyperspectral image datasets. The results show that the proposed method fully exploits the spectral–spatial features of hyperspectral images, and this can effectively improve the classification accuracy.

参考文献/References:

ZHANG M 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.
[2] LEE H, Kwon K. Going Deeper with Contextual CNN for Hyperspectral Image Classification[J]. IEEE Trans Image Process, 2017, 26(10): 4843-4855.
[3] WANG L, ZHANG J B, LIU P, et al. Spectral–spatial multi-feature-based deep learning for hyperspectral remote sensing image classification[J]. Soft Computing, 2017, 21(1): 213-221.
[4] Slavkovikj V, Verstockt S, Neve W D, et al. Unsupervised spectral sub-feature learning for hyperspectral image classification[J]. International Journal of Remote Sensing, 2016, 37(2): 309-326.
[5] HUANG H, ZHENG X L. Hyperspectral image classification with combination of weighted spatial-spectral and KNN[J]. Optics & Precision Engineering, 2016, 24(4): 873-881.
[6] 左航. 基于经验模态分解的SVM-KNN高光谱图像分类方法[J]. 微型电脑应用, 2016, 32(12): 60-63.
ZUO Hang. SVM-KNN Hyperspectral Image Classification Method Based on Empirical Mode Decomposition[J]. Microcomputer Application, 2016, 32(12): 60-63.
[7] HUANG K, LI S, KANG X, et al. Spectral–Spatial Hyperspectral Image Classification Based on KNN[J]. Sensing & Imaging, 2016, 17(1): 1-13.
[8] Echanobe J, Martinez M V, Basterretxea K, et al. Genetic Algorithm- based Optimization of ELM for On-line Hyperspectral Image Classification[C]//International Joint Conference on Neural Networks, 2017: 4202-4207.
[9] Jain D K, Dubey S B, Choubey R K,et al. An Approach for Hyperspectral Image Classification by Optimizing SVM using Self Organizing Map[J]. Journal of Computational Science, 2017, 25(1): 252-259.
[10] 李铁, 张新君. 基于联合协同表示与SVM决策融合的高光谱图像分类研究[J]. 计算机应用研究, 2017, 34(6): 1913-1916.
LI Tie, ZHANG Xinjun. Hyperspectral Image Classification Based on Joint Collaborative Representation and SVM Decision Fusion[J]. Journal of Computer Applications, 2017, 34(6): 1913-1916.
[11] Majdar R S, Ghassemian H. A probabilistic SVM approach for hyperspectral image classification using spectral and texture features[J]. International Journal of Remote Sensing, 2017, 38(15): 4265-4284.
[12] KANG X, XIANG X, LI S, et al. PCA-Based Edge-Preserving Features for Hyperspectral Image Classification[J]. IEEE Transactions on Geoscience & Remote Sensing, 2017, 55(12): 7140-7151.
[13] WANG W, ZHANG R, WU Q. Hyperspectral image classification based on PCA network[C]//Workshop on Hyperspectral Image & Signal Processing: Evolution in Remote Sensing, 2017: 1-4.
[14] Falco N, Benediktsson J A, Bruzzone L. A Study on the Effectiveness of Different Independent Component Analysis Algorithms for Hyperspectral Image Classification[J]. IEEE Journal of Selected Topics in Applied Earth Observations & Remote Sensing, 2014, 7(6): 2183-2199.
[15] LONG T, QIAN D, Kopriva I, et al. Tensor-based Offset-Sparsity Decomposition for Hyperspectral Image Classification[C]//Geoscience & Remote Sensing Symposium, 2017: 3656-3659.
[16] LI H C, ZHOU H L, PAN L, et al. Gabor Feature-based Composite Kernel Method for Hyperspectral Image Classification[J]. Electronics Letters, 2018, 54(10): 628-630.
[17] Mughees A, Ali A, TAO L M. Hyperspectral image classification via shape-adaptive deep learning[C]//IEEE International Conference on Image Processing, 2017: 375-379.
[18] WANG Q, LIN J Z, YUAN Y. Salient Band Selection for Hyperspectral Image Classification via Manifold Ranking[J]. IEEE Transactions on Neural Networks & Learning Systems, 2017, 27(6): 1279-1289.
[19] LI W, WU G D, ZHANG F, et al. Hyperspectral Image Classification Using Deep Pixel-Pair Features[J]. IEEE Transactions on Geoscience & Remote Sensing, 2016, 55(2): 844-853.
[20] LI Y, ZHANG H K, SHEN Q. Spectral-Spatial Classification of Hyperspectral Imagery with 3D Convolutional Neural Network[J]. Remote Sensing, 2017, 9(1): 1527-1554.
[21] Mou L, Ghamisi P,ZHU XX. Deep Recurrent Neural Networks for Hyperspectral Image Classification[J]. IEEE Transactions on Geoscience & Remote Sensing, 2017, 55(7): 3639-3655.
[22] WANG F, JIANG M Q, QIAN C, et al. Residual Attention Network for Image Classification[C]//IEEE Conference on Computer Vision and Pattern Recognition (CVPR), IEEE, 2017: 6450-6458.
[23] PENG Y X, HE X T, ZHAO J J. Object-Part Attention Model for Fine-grained Image Classification[J]. IEEE Transactions on Image Processing A Publication of the IEEE Signal Processing Society, 2018, 27(3): 1487-1500.
[24] Stanislau S, Erhardt B. Image classification with recurrent attention models[C]// Computational Intelligence, 2017:1-7.
[25] ZHONG Z L, LI J T, MA L F, et al. Deep Residual Networks for Hyperspectral Image Classification[C]//IEEE International Geoscience & Remote Sensing Symposium, 2017: 1824-1827.

备注/Memo

备注/Memo:
收稿日期:2019-07-12;修订日期:2020-02-29.
作者简介:王浩(1992-),男,硕士,主要从事高光谱图像分类的研究。E-mail: 386943738@qq.com。
通信作者:张晶晶(1974-),女,博士,副教授,主要从事偏振图像算法研究。E-mail:874878644@qq.com。
基金项目:安徽省自然科学基金(1808085MF209);偏振光成像探测技术安徽省重点实验室开放基金(2019KJS030009)。

更新日期/Last Update: 2020-03-17