基于局部保留降维卷积神经网络的高光谱图像分类算法
|
摘要
为提高高光谱遥感图像的分类精度,通过局部保留判别式分析与深度卷积神经网络(DCNN)算法,提出了基于局部保留降维卷积神经网络的高光谱图像分类算法。首先,用局部保留判别式分析对高光谱数据降维,再用二维Gabor滤波器对降维后的高光谱数据进行滤波,生成空间隧道信息;其次,用卷积神经网络对原始高光谱数据进行特征提取,生成光谱隧道信息;再次,融合空间隧道信息与光谱隧道信息,形成空间-光谱特征信息,并将其输入到深度卷积神经网络,提取更加有效的特征;最后,采用双重优选分类器对最终提取的特征进行分类。将本文方法与CNN、PCA-SVM、CD-CNN和CNN-PPF等算法在Indian Pines、University of Pavia高光谱遥感数据库上进行性能比较。在Indian Pines、University of Pavia数据库上,本文算法识别的整体精度比传统CNN方法的整体精度分别高3. 81个百分点与6. 62个百分点。实验结果表明,本文算法无论在分类精度还是Kappa系数都优于另外4种算法。
In order to improve the classification accuracy of hyperspectral remote sensing images,a novel hyperspectral image classification algorithm based on local preserving reduced dimensional convolutional neural network( DCNN) was proposed by using local preserving discriminant analysis and deep convolutional neural network( DCNN) algorithm. Firstly,the dimensionality reduction of hyperspectral data was analyzed by local reserved discriminant,and then the spatial tunnel information was filtered by two-dimensional Gabor filter. Secondly,the original hyperspectral data were extracted by convolution neural network to generate spectral tunnel information. Thirdly,the spatial tunnel information and spectral tunnel information were integrated to form the air-spectrum characteristic information,which was input into deep convolutional neural network to extract more effective features. Finally,the feature of the final extraction was classified by using the dual optimization classifier. The proposed method was compared with CNN,PCA-SVM,CD-CNN and CNN-PPF in the performance of Indian Pines and University of Pavia hyperspectral remote sensing databases. In the database of Indian Pines and University of Pavia,the overall recognition accuracy of the proposed method was 3. 81 percentage points and 6. 62 percentage points higher than that of the traditional CNN method. Experimental results on two databases showed that the proposed method was superior to the other four methods in both classification accuracy and Kappa coefficient,and it was a better classification method for hyperspectral remote sensing data classification.
In order to improve the classification accuracy of hyperspectral remote sensing images,a novel hyperspectral image classification algorithm based on local preserving reduced dimensional convolutional neural network( DCNN) was proposed by using local preserving discriminant analysis and deep convolutional neural network( DCNN) algorithm. Firstly,the dimensionality reduction of hyperspectral data was analyzed by local reserved discriminant,and then the spatial tunnel information was filtered by two-dimensional Gabor filter. Secondly,the original hyperspectral data were extracted by convolution neural network to generate spectral tunnel information. Thirdly,the spatial tunnel information and spectral tunnel information were integrated to form the air-spectrum characteristic information,which was input into deep convolutional neural network to extract more effective features. Finally,the feature of the final extraction was classified by using the dual optimization classifier. The proposed method was compared with CNN,PCA-SVM,CD-CNN and CNN-PPF in the performance of Indian Pines and University of Pavia hyperspectral remote sensing databases. In the database of Indian Pines and University of Pavia,the overall recognition accuracy of the proposed method was 3. 81 percentage points and 6. 62 percentage points higher than that of the traditional CNN method. Experimental results on two databases showed that the proposed method was superior to the other four methods in both classification accuracy and Kappa coefficient,and it was a better classification method for hyperspectral remote sensing data classification.
引文
[1] DU B,ZHANG Y,ZHANG L,et al. A hypothesis independent subpixel target detector for hyperspectral images[J]. Signal Processing,2015,110:244-249.
[2] YANG Sai,ZHU Qibing,HUANG Min. Application of joint skewness algorithm to select optimal wavelengths of hyperspectral image for maize seed classification[J]. Spectroscopy and Spectral Analysis,2017,37(3):990-996.
[3] DALM M,BUXTON M W N,RUITENBEEK F J A V. Discriminating ore and waste in a porphyry copper deposit using shortwavelength infrared(SWIR)hyperspectral imagery[J]. Minerals Engineering,2017,105:10-18.
[4] WASHBURN K E,STORMO S K,SKJELVAREID M H,et al. Non-invasive assessment of packaged cod freeze-thaw history by hyperspectral imaging[J]. Journal of Food Engineering,2017,205:64-73.
[5] LUO S,WANG C,XI X,et al. Fusion of airborne Li DAR data and hyperspectral imagery for aboveground and belowground forest biomass estimation[J]. Ecological Indicators,2017,73:378-387.
[6] XIA J,FALCO N,BENEDIKTSSON J A,et al. Hyperspectral image classification with rotation random forest via KPCA[J].IEEE Journal of Selected Topics in Applied Earth Observations&Remote Sensing,2017,10(4):1601-1609.
[7]王瀛,郭雷,梁楠.基于优选样本的KPCA高光谱图像降维方法[J].光子学报,2011,40(6):847-851.WANG Ying,GUO Lei,LIANG Nan. A dimensionality reduction method based on KPCA with optimized sample set for hyperspectral image[J]. Acta Photonica Sinica,2011,40(6):847-851.(in Chinese)
[8] ZHANG L,ZHANG L,DU B. Deep learning for remote sensing data:a technical tutorial on the sate of the art[J]. IEEE Geoscience&Remote Sensing Magazine,2016,4(2):22-40.
[9] ZHAO W,DU S. Spectral-spatial feature extraction for hyperspectral image classification:a dimension reduction and deep learning approach[J]. IEEE Transactions on Geoscience&Remote Sensing,2016,54(8):4544-4554.
[10] YUE J,ZHAO W,MAO S,et al. Spectral-spatial classification of hyperspectral images using deep convolutional neural networks[J]. Remote Sensing Letters,2015,6(6):468-477.
[11] MAKANTASIS K,KARANTZALOS K,DOULAMIS A,et al. Deep supervised learning for hyperspectral data classification through convolutional neural networks[C]∥Geoscience and Remote Sensing Symposium,2015:4959-4962.
[12] LI J,BIOUCAS-DIAS J M,PLAZA A. Spectral-spatial hyperspectral image segmentation using subsapce multinomial logisticregres sion and Markov random fields[J]. IEEE Transactions on Geoscience and Remote Sensing,2012,50(3):809-823.
[13] YANG J,ZHAO Y,CHAN C W,et al. Hyperspectral image classification using two-channel deep convolutional neural network[C]∥Geoscience and Remote Sensing Symposium,2016:5079-5082.
[14] LIU Q,ZHOUF,HANG R,et al. Bidirectional-convolutional LSTM based spectral-spatial feature learning for hyperspectral image classification[J]. Remote Sensing,2017,9(12):1330.
[15] SERRANOA,DIEGO I M,CONDA C,et al. Recent advances in face biometrics with Gabor wavelets:a review[J]. Pattern Recognition Letters,2010,31(5):372-381.
[16] KIM N C,SO H J. Directional statistical Gabor features for texture classification[J]. Pattern Recognition Letters,2018,112:18-26.
[17] LI W,DU Q. Gabor-filtering-based nearest regularized subspace for hyperspectral image classification[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing,2014,7(4):1012-1022.
[18] CHEN Y,JIANG H,LI C,et al. Deep feature extraction and classification of hyperspectral images based on convolutional neural networks[J]. IEEE Transactions on Geoscience and Remote Sensing,2016,54(10):6232-6251.
[19] RAJADELL O,GARCIA-SEVILLA P,PLA F. Spectral-spatial pixel characterization using Gabor filters for hyperspectral image classification[J]. IEEE Geoscience&Remote Sensing Letters,2013,10(4):860-864.
[20] LI W,TRAMEL E W,PRASAD S,et al. Nearest regularized subspace for hyperspectral classification[J]. IEEE Transactions on Geoscience&Remote Sensing,2013,52(1):477-489.
[21] KOETZ B,MORSDORF F,VAN DER L,et al. Multi-source land cover classification for forest fire management based on imaging spectrometry and Li DAR data[J]. Forest Ecology&Management,2008,256(3):263-271.
[22] HEIDEN U,HELDENS W,ROESSNER S,et al. Urban structure type characterization using hyperspectral remote sensing and height information[J]. Landscape&Urban Planning,2010,98(2):361-375.
[23]廖建尚,王立国,郝思嫒.基于双边滤波和空间邻域信息的高光谱图像分类方法[J/OL].农业机械学报,2017,48(8):140-146,211.LIAO Jianshang,WANG Liguo,HAO Siyuan. Hyperspectral image classification method combined with bilateral filtering and pixel neighborhood information[J/OL]. Transactions of the Chinese Society for Agricultural Machinery,2017,48(8):140-146,211. http:∥www. j-csam. org/jcsam/ch/reader/view_abstract. aspx? file_no=20170815&flag=1. DOI:10. 6041/j. issn.1000-1298. 2017. 08. 015.(in Chinese)
[24] BLACKBURN G A. Remote sensing of forest pigments using airborne imaging spectrometer and Li DAR imagery[J]. Remote Sensing of Environment,2002,82(2):311-321.
[25] ZHANG M,LI W,DU Q. Diverse region-based CNN for hyperspectral image classification[J]. IEEE Transactions on Image Processing,2018,27(6):2623-2634.
[2] YANG Sai,ZHU Qibing,HUANG Min. Application of joint skewness algorithm to select optimal wavelengths of hyperspectral image for maize seed classification[J]. Spectroscopy and Spectral Analysis,2017,37(3):990-996.
[3] DALM M,BUXTON M W N,RUITENBEEK F J A V. Discriminating ore and waste in a porphyry copper deposit using shortwavelength infrared(SWIR)hyperspectral imagery[J]. Minerals Engineering,2017,105:10-18.
[4] WASHBURN K E,STORMO S K,SKJELVAREID M H,et al. Non-invasive assessment of packaged cod freeze-thaw history by hyperspectral imaging[J]. Journal of Food Engineering,2017,205:64-73.
[5] LUO S,WANG C,XI X,et al. Fusion of airborne Li DAR data and hyperspectral imagery for aboveground and belowground forest biomass estimation[J]. Ecological Indicators,2017,73:378-387.
[6] XIA J,FALCO N,BENEDIKTSSON J A,et al. Hyperspectral image classification with rotation random forest via KPCA[J].IEEE Journal of Selected Topics in Applied Earth Observations&Remote Sensing,2017,10(4):1601-1609.
[7]王瀛,郭雷,梁楠.基于优选样本的KPCA高光谱图像降维方法[J].光子学报,2011,40(6):847-851.WANG Ying,GUO Lei,LIANG Nan. A dimensionality reduction method based on KPCA with optimized sample set for hyperspectral image[J]. Acta Photonica Sinica,2011,40(6):847-851.(in Chinese)
[8] ZHANG L,ZHANG L,DU B. Deep learning for remote sensing data:a technical tutorial on the sate of the art[J]. IEEE Geoscience&Remote Sensing Magazine,2016,4(2):22-40.
[9] ZHAO W,DU S. Spectral-spatial feature extraction for hyperspectral image classification:a dimension reduction and deep learning approach[J]. IEEE Transactions on Geoscience&Remote Sensing,2016,54(8):4544-4554.
[10] YUE J,ZHAO W,MAO S,et al. Spectral-spatial classification of hyperspectral images using deep convolutional neural networks[J]. Remote Sensing Letters,2015,6(6):468-477.
[11] MAKANTASIS K,KARANTZALOS K,DOULAMIS A,et al. Deep supervised learning for hyperspectral data classification through convolutional neural networks[C]∥Geoscience and Remote Sensing Symposium,2015:4959-4962.
[12] LI J,BIOUCAS-DIAS J M,PLAZA A. Spectral-spatial hyperspectral image segmentation using subsapce multinomial logisticregres sion and Markov random fields[J]. IEEE Transactions on Geoscience and Remote Sensing,2012,50(3):809-823.
[13] YANG J,ZHAO Y,CHAN C W,et al. Hyperspectral image classification using two-channel deep convolutional neural network[C]∥Geoscience and Remote Sensing Symposium,2016:5079-5082.
[14] LIU Q,ZHOUF,HANG R,et al. Bidirectional-convolutional LSTM based spectral-spatial feature learning for hyperspectral image classification[J]. Remote Sensing,2017,9(12):1330.
[15] SERRANOA,DIEGO I M,CONDA C,et al. Recent advances in face biometrics with Gabor wavelets:a review[J]. Pattern Recognition Letters,2010,31(5):372-381.
[16] KIM N C,SO H J. Directional statistical Gabor features for texture classification[J]. Pattern Recognition Letters,2018,112:18-26.
[17] LI W,DU Q. Gabor-filtering-based nearest regularized subspace for hyperspectral image classification[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing,2014,7(4):1012-1022.
[18] CHEN Y,JIANG H,LI C,et al. Deep feature extraction and classification of hyperspectral images based on convolutional neural networks[J]. IEEE Transactions on Geoscience and Remote Sensing,2016,54(10):6232-6251.
[19] RAJADELL O,GARCIA-SEVILLA P,PLA F. Spectral-spatial pixel characterization using Gabor filters for hyperspectral image classification[J]. IEEE Geoscience&Remote Sensing Letters,2013,10(4):860-864.
[20] LI W,TRAMEL E W,PRASAD S,et al. Nearest regularized subspace for hyperspectral classification[J]. IEEE Transactions on Geoscience&Remote Sensing,2013,52(1):477-489.
[21] KOETZ B,MORSDORF F,VAN DER L,et al. Multi-source land cover classification for forest fire management based on imaging spectrometry and Li DAR data[J]. Forest Ecology&Management,2008,256(3):263-271.
[22] HEIDEN U,HELDENS W,ROESSNER S,et al. Urban structure type characterization using hyperspectral remote sensing and height information[J]. Landscape&Urban Planning,2010,98(2):361-375.
[23]廖建尚,王立国,郝思嫒.基于双边滤波和空间邻域信息的高光谱图像分类方法[J/OL].农业机械学报,2017,48(8):140-146,211.LIAO Jianshang,WANG Liguo,HAO Siyuan. Hyperspectral image classification method combined with bilateral filtering and pixel neighborhood information[J/OL]. Transactions of the Chinese Society for Agricultural Machinery,2017,48(8):140-146,211. http:∥www. j-csam. org/jcsam/ch/reader/view_abstract. aspx? file_no=20170815&flag=1. DOI:10. 6041/j. issn.1000-1298. 2017. 08. 015.(in Chinese)
[24] BLACKBURN G A. Remote sensing of forest pigments using airborne imaging spectrometer and Li DAR imagery[J]. Remote Sensing of Environment,2002,82(2):311-321.
[25] ZHANG M,LI W,DU Q. Diverse region-based CNN for hyperspectral image classification[J]. IEEE Transactions on Image Processing,2018,27(6):2623-2634.