基于卷积神经网络的低空风切变类型识别
|
摘要
针对激光雷达低空风切变信号图像的类型识别问题,提出了一种基于深度卷积神经网络(deep convolutional neural network,DCNN)的多层特征提取及自适应融合算法。该方法可以有效解决网络逐层训练过程中信息丢失的问题。首先,采用DCNN提取低空风切变信号图像的各层网络特征,并将各特征进行L2范数标准化实现同趋化。其次,将其以多通道图像形式输入单层CNN进行自适应融合,将融合特征送入支持向量机进行分类识别。结果表明,采用所提算法进行低空风切变图像类型识别的平均识别率为98.1%,与其他4种算法相比均有提升。所提算法能更有效地实现低空风切变信号图像类型识别。
This paper proposes an algorithm for multilayer feature extraction and adaptive fusion based on deep convolutional neural network(DCNN)to solve the problem of image type recognition of the low-level wind shear signal scanned by the laser radar.The method effectively makes up for the information loss in the process of layer-by-layer network training.First of all,the characteristics of the network layers of low-level wind shear signal image are extracted by using DCNN modeling and the characteristics are L2-norm standardized.Then the L2-norm standardized characteristics are put into single-layer CNN for adaptive fusion in the form of multichannel image and the fusion characeristics are sent into the support vector machine for classification recognition.The results show that the average recognition rate of image type recognition of the low-level wind shear signal obtained by using the proposed algorithm is 98.1%,being improved compared with the other four algorithms.The algorithm can effectively realize image type recognition of the low-level wind shear signal.
This paper proposes an algorithm for multilayer feature extraction and adaptive fusion based on deep convolutional neural network(DCNN)to solve the problem of image type recognition of the low-level wind shear signal scanned by the laser radar.The method effectively makes up for the information loss in the process of layer-by-layer network training.First of all,the characteristics of the network layers of low-level wind shear signal image are extracted by using DCNN modeling and the characteristics are L2-norm standardized.Then the L2-norm standardized characteristics are put into single-layer CNN for adaptive fusion in the form of multichannel image and the fusion characeristics are sent into the support vector machine for classification recognition.The results show that the average recognition rate of image type recognition of the low-level wind shear signal obtained by using the proposed algorithm is 98.1%,being improved compared with the other four algorithms.The algorithm can effectively realize image type recognition of the low-level wind shear signal.
引文
[1]MA Y,LI S,LU W.Recognition of partial scanning low-level wind shear based on support vector machine[J].Advances in Mechanical Engineering,2018,10(1):1-7.
[2]TIANA-ALSINA J,ROCADENBOSCH F,GUTIERREZ-ANTUNANO M A.Vertical azimuth display simulator for wind-Doppler lidar error assessment[C]∥Proc.of the Geoscience and Remote Sensing Symposium,2017:1614-1617.
[3]蒋立辉,田百全,熊兴隆,等.基于多普勒激光雷达低空风切变的数值仿真[J].红外与激光工程,2012,41(7):1761-1766.JIANG L H,TIAN B Q,XIONG X L,et al.Numerical simulations of low attitude wind shear based on Doppler lidar[J].Infrared and Laser Engineering,2012,41(7):1761-1766.
[4]ZHOU Y,ZHANG R,WANG S,et al.Feature selection method based on high-resolution remote sensing images and the effect of sensitive features on classification accuracy[J].Sensors,2018,18(7):E2013.
[5]蒋立辉,高志光,熊兴隆,等.基于激光雷达图像处理的低空风切变类型识别研究[J].红外与激光工程,2012,41(12):3410-3415.JIANG L H,GAO Z G,XIONG X L,et al.Study on type recognition of low attitude wind shear based on lidar image processing[J].Infrared and Laser Engineering,2012,41(12):3410-3415.
[6]ZHANG R,ZHANG M.SAR target recognition based on active contour without edges[J].Journal of Systems Engineering and Electronics,2017,28(2):276-281.
[7]SUCIATI N,HERUMURTI D,WIJAYA A Y.Feature extraction using gray-level co-occurrence matrix of wavelet coefficients and texture matching for batik motif recognition[J].Proceedings of the SPIE,2017,225:1022502.
[8]蒋立辉,高浩,庄子波,等.激光雷达中基于组合纹理的低空风切变识别算法[J].红外与激光,2014,44(3):325-329.JIANG L H,GAO H,ZHUANG Z B,et al.Recognition algorithm of low altitude wind shear based on combination texture[J].Laser&Infrared,2014,44(3):325-329.
[9]朱天呈,黄世震.基于分块加权的LBP算法在人脸识别中的应用[J].河南科技,2018,643(8):12-14.ZHU T X,HUANG S Z.Application of LBP algorithm based on block weighting in face recognition[J].Henan Science&Technology,2018,643(8):12-14.
[10]UBUL K,YADIKAR N,AMAT A,et al.Uyghur document image retrieval based on gray gradient co-occurrence matrix[C]∥Proc.of the Chinese Automation Congress,2016:762-766.
[11]于明,安梦涛,刘依.基于多特征与卷积神经网络的人脸表情识别[J].科学技术与工程,2018,18(13):104-110.YU M,AN M T,LIU Y.Facial expression recognition based on multiple features and convolutional neural networks[J].Science Technology and Engineering,2018,18(13):104-110.
[12]ZHOU J X,LIU X D,XU T W,et al.A new fusion approach for content based image retrieval with color histogram and local directional pattern[J].International Journal of Machine Learning&Cybernetics,2018,9(4):677-689.
[13]GIRI K J,BASHIR R.A block based watermarking approach for color images using discrete wavelet transformation[J].International Journal of Information Technology,2017(3):1-8.
[14]化春健,熊雪梅,陈莹.基于Sobel算子的工件圆弧轮廓特征提取[J].激光与光电子学进展,2018,55(2):21011-1.HUA C J,XIONG X M,CHEN Y.Feature extraction of workpiece circular arc contour based on sobel operator[J].Laser&Optoelectronics Progress,2018,55(2):21011-1.
[15]王念滨,何鸣,王红滨,等.基于卷积神经网络的水下目标特征提取方法[J].系统工程与电子技术,2018,40(6):1197-1203.WANG N B,HE M,WANG H B,et al.Underwater target feature extraction method based on convolutional neural network[J].Systems Engineering and Electronics,2018,40(6):1197-1203.
[16]VERMA A,VIG L.Using convolutional neural networks to discover cogntively validated features for gender classification[C]∥Proc.of the IEEE International Conference on Soft Computing and Machine Intelligence,2015:33-37.
[17]黄旭,凌志刚,李绣心.融合判别式深度特征学习的图像识别算法[J].中国图象图形学报,2018,23(4):510-518.HUANG X,LING Z G,LI X X.Discriminative deep feature learning method by fusing linear discriminant analysis for image recognition[J].Journal of Image and Graphics,2018,23(4):510-518.
[18]SHEN W,WANG X,WANG Y,et al.Deep contour:a deep convolutional feature learned by positive-sharing loss for contour detection[C]∥Proc.of the IEEE Computer Vision and Pattern Recognition,2015:3982-3991.
[19]罗畅,王洁,王世强,等.基于泛化深度迁移特征的高分遥感场景分类[J].系统工程与电子技术,2018,40(3):682-691.LUO C,WANG J,WANG S Q,et al.General deep transfer features based high resolution remote scene classification[J].Systems Engineering and Electronics,2018,40(3):682-691.
[20]RANJAN R,CASTILLO C D,CHELLAPPA R.L2-constrained softmax loss for discriminative face verification[J].Computer Vision and Pattern Recognition,2017.
[21]NIU X X,SUEN C Y.A novel hybrid CNN-SVM classifier for recognizing handwritten digits[J].Pattern Recognition,2011,45(4):1318-1325.
[22]WU Y,JIANG M.Face recognition system based on CNN and LBP features for classifier optimization and fusion[J].The Journal of China University of Posts and Telecommunications,2018,25(1):37-47.
[2]TIANA-ALSINA J,ROCADENBOSCH F,GUTIERREZ-ANTUNANO M A.Vertical azimuth display simulator for wind-Doppler lidar error assessment[C]∥Proc.of the Geoscience and Remote Sensing Symposium,2017:1614-1617.
[3]蒋立辉,田百全,熊兴隆,等.基于多普勒激光雷达低空风切变的数值仿真[J].红外与激光工程,2012,41(7):1761-1766.JIANG L H,TIAN B Q,XIONG X L,et al.Numerical simulations of low attitude wind shear based on Doppler lidar[J].Infrared and Laser Engineering,2012,41(7):1761-1766.
[4]ZHOU Y,ZHANG R,WANG S,et al.Feature selection method based on high-resolution remote sensing images and the effect of sensitive features on classification accuracy[J].Sensors,2018,18(7):E2013.
[5]蒋立辉,高志光,熊兴隆,等.基于激光雷达图像处理的低空风切变类型识别研究[J].红外与激光工程,2012,41(12):3410-3415.JIANG L H,GAO Z G,XIONG X L,et al.Study on type recognition of low attitude wind shear based on lidar image processing[J].Infrared and Laser Engineering,2012,41(12):3410-3415.
[6]ZHANG R,ZHANG M.SAR target recognition based on active contour without edges[J].Journal of Systems Engineering and Electronics,2017,28(2):276-281.
[7]SUCIATI N,HERUMURTI D,WIJAYA A Y.Feature extraction using gray-level co-occurrence matrix of wavelet coefficients and texture matching for batik motif recognition[J].Proceedings of the SPIE,2017,225:1022502.
[8]蒋立辉,高浩,庄子波,等.激光雷达中基于组合纹理的低空风切变识别算法[J].红外与激光,2014,44(3):325-329.JIANG L H,GAO H,ZHUANG Z B,et al.Recognition algorithm of low altitude wind shear based on combination texture[J].Laser&Infrared,2014,44(3):325-329.
[9]朱天呈,黄世震.基于分块加权的LBP算法在人脸识别中的应用[J].河南科技,2018,643(8):12-14.ZHU T X,HUANG S Z.Application of LBP algorithm based on block weighting in face recognition[J].Henan Science&Technology,2018,643(8):12-14.
[10]UBUL K,YADIKAR N,AMAT A,et al.Uyghur document image retrieval based on gray gradient co-occurrence matrix[C]∥Proc.of the Chinese Automation Congress,2016:762-766.
[11]于明,安梦涛,刘依.基于多特征与卷积神经网络的人脸表情识别[J].科学技术与工程,2018,18(13):104-110.YU M,AN M T,LIU Y.Facial expression recognition based on multiple features and convolutional neural networks[J].Science Technology and Engineering,2018,18(13):104-110.
[12]ZHOU J X,LIU X D,XU T W,et al.A new fusion approach for content based image retrieval with color histogram and local directional pattern[J].International Journal of Machine Learning&Cybernetics,2018,9(4):677-689.
[13]GIRI K J,BASHIR R.A block based watermarking approach for color images using discrete wavelet transformation[J].International Journal of Information Technology,2017(3):1-8.
[14]化春健,熊雪梅,陈莹.基于Sobel算子的工件圆弧轮廓特征提取[J].激光与光电子学进展,2018,55(2):21011-1.HUA C J,XIONG X M,CHEN Y.Feature extraction of workpiece circular arc contour based on sobel operator[J].Laser&Optoelectronics Progress,2018,55(2):21011-1.
[15]王念滨,何鸣,王红滨,等.基于卷积神经网络的水下目标特征提取方法[J].系统工程与电子技术,2018,40(6):1197-1203.WANG N B,HE M,WANG H B,et al.Underwater target feature extraction method based on convolutional neural network[J].Systems Engineering and Electronics,2018,40(6):1197-1203.
[16]VERMA A,VIG L.Using convolutional neural networks to discover cogntively validated features for gender classification[C]∥Proc.of the IEEE International Conference on Soft Computing and Machine Intelligence,2015:33-37.
[17]黄旭,凌志刚,李绣心.融合判别式深度特征学习的图像识别算法[J].中国图象图形学报,2018,23(4):510-518.HUANG X,LING Z G,LI X X.Discriminative deep feature learning method by fusing linear discriminant analysis for image recognition[J].Journal of Image and Graphics,2018,23(4):510-518.
[18]SHEN W,WANG X,WANG Y,et al.Deep contour:a deep convolutional feature learned by positive-sharing loss for contour detection[C]∥Proc.of the IEEE Computer Vision and Pattern Recognition,2015:3982-3991.
[19]罗畅,王洁,王世强,等.基于泛化深度迁移特征的高分遥感场景分类[J].系统工程与电子技术,2018,40(3):682-691.LUO C,WANG J,WANG S Q,et al.General deep transfer features based high resolution remote scene classification[J].Systems Engineering and Electronics,2018,40(3):682-691.
[20]RANJAN R,CASTILLO C D,CHELLAPPA R.L2-constrained softmax loss for discriminative face verification[J].Computer Vision and Pattern Recognition,2017.
[21]NIU X X,SUEN C Y.A novel hybrid CNN-SVM classifier for recognizing handwritten digits[J].Pattern Recognition,2011,45(4):1318-1325.
[22]WU Y,JIANG M.Face recognition system based on CNN and LBP features for classifier optimization and fusion[J].The Journal of China University of Posts and Telecommunications,2018,25(1):37-47.