卷积神经网络特征在遥感图像配准中的应用
|
摘要
遥感图像配准是许多遥感应用的重要步骤之一。卷积神经网络(convolutional neural network,CNN)提取的图像高层特征在图像分类和检索问题上表现优异,能够克服低层配准特征的表达能力有限、容易受到干扰等问题。因此对利用CNN特征进行遥感图像配准开展研究。首先,针对遥感图像配准问题,对CNN中的全连接层特征和不同聚合大小的卷积层特征进行了研究;然后,对利用CNN特征进行图像配准的方法进行了分析;最后,将CNN特征与尺度不变特征变换(scale-invariant feature transform,SIFT)特征在图像的旋转角度、缩放倍数和亮度依次变换时的配准性能进行了对比分析。实验结果表明,在匹配精度和正确对应点的数量方面,CNN特征比SIFT方法具有更好的匹配性能;对变换后的图像而言,微调后的CNN特征比SIFT特征具有更强的鲁棒性。
Successful remote sensing image registration is one of the foundations of many remote sensing applications. Image high-lever features extracted by convolutional neural network( CNN) have achieved excellent performance in image classification and retrieval,and can be used to solve some problems of low-lever image registration features,such as the limitation of expression capability and easily being interfered. Hence,in this paper,the authors investigated the problem as to how to use CNN feature for remote sensing image registration.First,the authors investigated different CNN features from fully connected layers and aggregating convolutional features with different sizes from convolutional layer to register remote sensing image. Then the authors introduced the procedure by using CNN feature for image registration. Finally, the authors compared the registration performances of CNN features and scale-invariant feature transform( SIFT) features after the transformation of the image's perspective,brightness and scale,respectively. The experimental results show that the CNN feature has better matching performance than the SIFT method in terms of matching accuracy and correct number of corresponding points. The finely tuned CNN feature has stronger robustness to the transformed image than the SIFT feature.
Successful remote sensing image registration is one of the foundations of many remote sensing applications. Image high-lever features extracted by convolutional neural network( CNN) have achieved excellent performance in image classification and retrieval,and can be used to solve some problems of low-lever image registration features,such as the limitation of expression capability and easily being interfered. Hence,in this paper,the authors investigated the problem as to how to use CNN feature for remote sensing image registration.First,the authors investigated different CNN features from fully connected layers and aggregating convolutional features with different sizes from convolutional layer to register remote sensing image. Then the authors introduced the procedure by using CNN feature for image registration. Finally, the authors compared the registration performances of CNN features and scale-invariant feature transform( SIFT) features after the transformation of the image's perspective,brightness and scale,respectively. The experimental results show that the CNN feature has better matching performance than the SIFT method in terms of matching accuracy and correct number of corresponding points. The finely tuned CNN feature has stronger robustness to the transformed image than the SIFT feature.
引文
[1] Zitova B,Flusser J. Image registration methods:A survey[J]. Image and Vision Computing,2003,21(11):977-1000.
[2] Gong M,Zhao S,Jiao L,et al. A novel coarse-to-ne scheme for automatic image registration based on sift and mutual information[J]. IEEE Transactions on Geoscience and Remote Sensing,2014,52(7):4328-4338.
[3]张谦,贾永红,胡忠文.多源遥感影像配准中的SIFT特征匹配改进[J].武汉大学学报(信息科学版),2013,38(4):455-459.Zhang Q,Jia Y H,Hu Z W. An improved SIFT algorithm for multisource remote sensing image registration[J]. Geomatics and Information Science of Wuhan University,2013,38(4):455-459.
[4]李少毅,王晓田,杨开.改进的SURF彩色遥感图像配准算法[J].计算机测量与控制,2017,25(1):209-212.Li S Y,Wang X T,Yang K. An improved SURF algorithm for color remote sensing image registration[J]. Computer Measurement and Control,2017,25(1):209-212.
[5] Yang K,Karlstrom L,Smith L C,et al. Automated high-resolution satellite image registration using supraglacial rivers on the Greenland ice sheet[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing,2017,99:1-12.
[6] Krizhevsky A,Sutskever I,Hinton G E. Image Net classification with deep convolutional neural networks[J]. Advance in Neural Information Processing Systems,2012,25(2):1097-1105.
[7] Simonyan K,Zisserman A. Very deep convolutional networks for largescale image recognition[EB/OL].(2015-04-10). http://arxiv. org/pdf/1409. 1556. pdf.
[8] Chandrasekhar V,Lin J,Morère O,et al. A practical guide to CNNs and fisher vectors for image instance retrieval[J]. Signal Processing,2015,128:426-439.
[9]罗建豪,吴建鑫.基于深度卷积特征的细粒度图像分类研究综述[J].自动化学报,2017,43(8):1306-1318.Luo J H,Wu J X. A survey on fine-grained image categorization using deep convolutional features[J]. Acta Automatica Sinica,2017,43(8):1306-1318.
[10]张洪群,刘雪莹,杨森,等.深度学习的半监督遥感图像检索[J].遥感学报,2017,21(3):406-414.Zhang H Q,Liu X Y,Yang S,et al. Retrieval of remote sensing images based on semisupervised deep learning[J]. Journal of Remote Sensing,2017,21(3):406-414.
[11]刘峰,沈同圣,马新星.特征融合的卷积神经网络多波段舰船目标识别[J].光学学报,2017,37(10):248-256.Liu F,Shen T S,Ma X X. Convolutional neural network based multi-band ship target recognition with feature fusion[J]. Acta Optica Sinica,2017,37(10):248-256.
[12] Zhu G,Wang Q,Yuan Y,et al. SIFT on manifold:An intrinsic description[J]. Neurocomputing,2013,113(7):227-233.
[13] Yosinski J,Clune J,Bengio Y,et al. How transferable are features in deep neural networks?[C]//International Conference on Neural Information Processing Systems. MIT Press,2014:3320-3328.
[14] Babenko A,Lempitsky V. Aggregating deep convolutional features for image retrieval[EB/OL].(2015-10-26). http://arxiv.org/pdf/1510. 07493v1. pdf.
[15] Wei X S,Luo J H,Wu J,et al. Selective convolutional descriptor aggregation for fine-grained image retrieval[J]. IEEE Transactions on Image Processing,2017,26(6):2868-2881.
[16] Goncalves H,Goncalves J A,Corte-Real L. Measures for an objective evaluation of the geometric correction process quality[J].IEEE Geoscience and Remote Sensing Letters,2009,6(2):292-296.
[2] Gong M,Zhao S,Jiao L,et al. A novel coarse-to-ne scheme for automatic image registration based on sift and mutual information[J]. IEEE Transactions on Geoscience and Remote Sensing,2014,52(7):4328-4338.
[3]张谦,贾永红,胡忠文.多源遥感影像配准中的SIFT特征匹配改进[J].武汉大学学报(信息科学版),2013,38(4):455-459.Zhang Q,Jia Y H,Hu Z W. An improved SIFT algorithm for multisource remote sensing image registration[J]. Geomatics and Information Science of Wuhan University,2013,38(4):455-459.
[4]李少毅,王晓田,杨开.改进的SURF彩色遥感图像配准算法[J].计算机测量与控制,2017,25(1):209-212.Li S Y,Wang X T,Yang K. An improved SURF algorithm for color remote sensing image registration[J]. Computer Measurement and Control,2017,25(1):209-212.
[5] Yang K,Karlstrom L,Smith L C,et al. Automated high-resolution satellite image registration using supraglacial rivers on the Greenland ice sheet[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing,2017,99:1-12.
[6] Krizhevsky A,Sutskever I,Hinton G E. Image Net classification with deep convolutional neural networks[J]. Advance in Neural Information Processing Systems,2012,25(2):1097-1105.
[7] Simonyan K,Zisserman A. Very deep convolutional networks for largescale image recognition[EB/OL].(2015-04-10). http://arxiv. org/pdf/1409. 1556. pdf.
[8] Chandrasekhar V,Lin J,Morère O,et al. A practical guide to CNNs and fisher vectors for image instance retrieval[J]. Signal Processing,2015,128:426-439.
[9]罗建豪,吴建鑫.基于深度卷积特征的细粒度图像分类研究综述[J].自动化学报,2017,43(8):1306-1318.Luo J H,Wu J X. A survey on fine-grained image categorization using deep convolutional features[J]. Acta Automatica Sinica,2017,43(8):1306-1318.
[10]张洪群,刘雪莹,杨森,等.深度学习的半监督遥感图像检索[J].遥感学报,2017,21(3):406-414.Zhang H Q,Liu X Y,Yang S,et al. Retrieval of remote sensing images based on semisupervised deep learning[J]. Journal of Remote Sensing,2017,21(3):406-414.
[11]刘峰,沈同圣,马新星.特征融合的卷积神经网络多波段舰船目标识别[J].光学学报,2017,37(10):248-256.Liu F,Shen T S,Ma X X. Convolutional neural network based multi-band ship target recognition with feature fusion[J]. Acta Optica Sinica,2017,37(10):248-256.
[12] Zhu G,Wang Q,Yuan Y,et al. SIFT on manifold:An intrinsic description[J]. Neurocomputing,2013,113(7):227-233.
[13] Yosinski J,Clune J,Bengio Y,et al. How transferable are features in deep neural networks?[C]//International Conference on Neural Information Processing Systems. MIT Press,2014:3320-3328.
[14] Babenko A,Lempitsky V. Aggregating deep convolutional features for image retrieval[EB/OL].(2015-10-26). http://arxiv.org/pdf/1510. 07493v1. pdf.
[15] Wei X S,Luo J H,Wu J,et al. Selective convolutional descriptor aggregation for fine-grained image retrieval[J]. IEEE Transactions on Image Processing,2017,26(6):2868-2881.
[16] Goncalves H,Goncalves J A,Corte-Real L. Measures for an objective evaluation of the geometric correction process quality[J].IEEE Geoscience and Remote Sensing Letters,2009,6(2):292-296.