基于多层深度特征融合的极化合成孔径雷达图像语义分割
|
摘要
针对传统特征表征能力较弱的问题,提出了一种基于多层深度特征融合的极化合成孔径雷达图像语义分割方法;利用经过预训练的VGG-Net-16模型提取表征能力更强的多层图像特征,再将各层深度特征分别用于训练对应的条件随机场模型,最后将多个条件随机场模型的输出结果进行融合,实现了最终的图像语义分割。结果表明:与基于传统经典特征的方法相比,所提方法取得了最高的总体分类精度,说明所提方法采用的融合特征具有比传统特征更强的表征能力。
Aiming at the problem that the traditional feature representation ability is weak, we propose a polarization synthetic aperture radar image semantic segmentation method based on the multi-layer deep feature fusion. The pre-trained VGG-Net-16 model is used to extract multi-layer image features with strong representation ability, and then deep features of each layer are used to train the corresponding conditional random field model. The output results of multiple conditional random field models are finally merged to realize the final semantic segmentation of the images. The results show that compared with the methods based on classical features, the proposed method achieves the highest overall accuracy, indicating that the fusion features used by the proposed method have stronger representation ability than traditional features.
Aiming at the problem that the traditional feature representation ability is weak, we propose a polarization synthetic aperture radar image semantic segmentation method based on the multi-layer deep feature fusion. The pre-trained VGG-Net-16 model is used to extract multi-layer image features with strong representation ability, and then deep features of each layer are used to train the corresponding conditional random field model. The output results of multiple conditional random field models are finally merged to realize the final semantic segmentation of the images. The results show that compared with the methods based on classical features, the proposed method achieves the highest overall accuracy, indicating that the fusion features used by the proposed method have stronger representation ability than traditional features.
引文
[1] Yang W, Zhang X, Chen L J, et al. Semantic segmentation of polarimetric SAR imagery using conditional random fields[C]//IEEE International Geoscience and Remote Sensing Symposium, July 25-30, 2010, Honolulu, Hawaii, USA. New Jersey: IEEE, 2010: 1593-1596.
[2] Chen S W, Tao C S. PolSAR image classification using polarimetric-feature-driven deep convolutional neural network[J]. IEEE Geoscience and Remote Sensing Letters, 2018, 15(4): 627-631.
[3] Wan J H, Zang J X, Liu S W. Fusion and classification of SAR and optical image with consideration of polarization characteristics[J]. Acta Optica Sinica, 2017, 37(6): 0628001. 万剑华, 臧金霞, 刘善伟. 顾及极化特征的SAR与光学影像融合与分类[J]. 光学学报, 2017, 37(6): 0628001.
[4] Krizhevsky A, Sutskever I, Hinton G E. ImageNet classification with deep convolutional neural networks[J]. Communications of the ACM, 2017, 60(6): 84-90.
[5] Simonyan K, Zisserman A. Very deep convolutional networks for large-scale image recognition[J]. Computer Science, 2014, 32(74): 2965-2971.
[6] He K M, Zhang X Y, Ren S Q, et al. Delving deep into rectifiers: surpassing human-level performance on ImageNet classification[J]. Proceedings of the IEEE, 2015, 28: 1026-1034.
[7] Zhou Y, Wang H P, Xu F, et al. Polarimetric SAR image classification using deep convolutional neural networks [J]. IEEE Geoscience and Remote Sensing Letters, 2016, 13(12): 1935-1939.
[8] Hou B, Kou H D, Jiao L C. Classification of polarimetric SAR images using multilayer autoencoders and superpixels[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2016, 9(7): 3072-3081.
[9] Lafferty J, McCallum A, Pereira F. Conditional random fields: probabilistic models for segmenting and labeling sequence data[C]//Proceedings of the 18th International Conference on Machine Learning, Morgan Kaufmann, San Francisco, CA. San Francisco: Morgan Kaufmann, 2001: 282-289.
[10] LI S Z. Markov random field modeling in computer vision[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1995, 17(10): 939-954.
[11] Han P, Han B B. Lee filter of PolSAR image based on typical scattering difference index[J]. Systems Engineering and Electronics, 2018, 40(2): 287-294. 韩萍, 韩宾宾. 基于典型散射差异指数的PolSAR图像Lee滤波[J]. 系统工程与电子技术, 2018, 40(2): 287-294.
[12] Mika S, Sch?lkopf B, Smola A, et al. Kernel PCA and de-noising in feature spaces[C]//Proceedings of Conference on Advances in Neural Information Processing Systems. Cambridge: The MIT Press, 1999: 536-542.
[13] Lee C H, Schmidt M, Murtha A, et al. Segmenting brain tumors with conditional random fields and support vector machines[C]//Proceedings of the 1st Internatinoal Conference on Computer Vision for Biomedical Image Applications. Heidelberg: Springer, 2005: 469-478.
[14] Kumar S, Hebert M. Discriminative fields for modeling spatial dependencies in natural images[C]//Proceedings of Conference on Advances in Neural Information Processing Systems. Cambridge: The MIT Press, 2004: 1531-1538.
[15] Domke J. Learning graphical model parameters with approximate marginal inference[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2013, 35(10): 2454-2467.
[16] Parikh D, Batra D. CRFs for image classification[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2000, 22(5): 460-472.
[17] Krishnapuram B, Carin L, Figueiredo M, et al. Sparse multinomial logistic regression: fast algorithms and generalization bounds[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2005, 27(6): 957-968.
[18] Vedaldi A, Lenc K. MatConvNet: convolutional neural networks for MATLAB[C]//Proceedings of the 23rd ACM International Conference on Multimedia, October 26-30, 2015, Brisbane, Australia. New York: ACM, 2015: 689-692.
[19] Yu P, Qin A K, Clausi D A. Unsupervised polarimetric SAR image segmentation and classification using region growing with edge penalty[J]. IEEE Transactions on Geoscience and Remote Sensing, 2011, 50(4): 1302-1317.
[20] Liu B, Hu H, Wang H Y, et al. Superpixel-based classification with an adaptive number of classes for polarimetric SAR images[J]. IEEE Transactions on Geoscience and Remote Sensing, 2013, 51(2): 907-924.
[2] Chen S W, Tao C S. PolSAR image classification using polarimetric-feature-driven deep convolutional neural network[J]. IEEE Geoscience and Remote Sensing Letters, 2018, 15(4): 627-631.
[3] Wan J H, Zang J X, Liu S W. Fusion and classification of SAR and optical image with consideration of polarization characteristics[J]. Acta Optica Sinica, 2017, 37(6): 0628001. 万剑华, 臧金霞, 刘善伟. 顾及极化特征的SAR与光学影像融合与分类[J]. 光学学报, 2017, 37(6): 0628001.
[4] Krizhevsky A, Sutskever I, Hinton G E. ImageNet classification with deep convolutional neural networks[J]. Communications of the ACM, 2017, 60(6): 84-90.
[5] Simonyan K, Zisserman A. Very deep convolutional networks for large-scale image recognition[J]. Computer Science, 2014, 32(74): 2965-2971.
[6] He K M, Zhang X Y, Ren S Q, et al. Delving deep into rectifiers: surpassing human-level performance on ImageNet classification[J]. Proceedings of the IEEE, 2015, 28: 1026-1034.
[7] Zhou Y, Wang H P, Xu F, et al. Polarimetric SAR image classification using deep convolutional neural networks [J]. IEEE Geoscience and Remote Sensing Letters, 2016, 13(12): 1935-1939.
[8] Hou B, Kou H D, Jiao L C. Classification of polarimetric SAR images using multilayer autoencoders and superpixels[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2016, 9(7): 3072-3081.
[9] Lafferty J, McCallum A, Pereira F. Conditional random fields: probabilistic models for segmenting and labeling sequence data[C]//Proceedings of the 18th International Conference on Machine Learning, Morgan Kaufmann, San Francisco, CA. San Francisco: Morgan Kaufmann, 2001: 282-289.
[10] LI S Z. Markov random field modeling in computer vision[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1995, 17(10): 939-954.
[11] Han P, Han B B. Lee filter of PolSAR image based on typical scattering difference index[J]. Systems Engineering and Electronics, 2018, 40(2): 287-294. 韩萍, 韩宾宾. 基于典型散射差异指数的PolSAR图像Lee滤波[J]. 系统工程与电子技术, 2018, 40(2): 287-294.
[12] Mika S, Sch?lkopf B, Smola A, et al. Kernel PCA and de-noising in feature spaces[C]//Proceedings of Conference on Advances in Neural Information Processing Systems. Cambridge: The MIT Press, 1999: 536-542.
[13] Lee C H, Schmidt M, Murtha A, et al. Segmenting brain tumors with conditional random fields and support vector machines[C]//Proceedings of the 1st Internatinoal Conference on Computer Vision for Biomedical Image Applications. Heidelberg: Springer, 2005: 469-478.
[14] Kumar S, Hebert M. Discriminative fields for modeling spatial dependencies in natural images[C]//Proceedings of Conference on Advances in Neural Information Processing Systems. Cambridge: The MIT Press, 2004: 1531-1538.
[15] Domke J. Learning graphical model parameters with approximate marginal inference[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2013, 35(10): 2454-2467.
[16] Parikh D, Batra D. CRFs for image classification[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2000, 22(5): 460-472.
[17] Krishnapuram B, Carin L, Figueiredo M, et al. Sparse multinomial logistic regression: fast algorithms and generalization bounds[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2005, 27(6): 957-968.
[18] Vedaldi A, Lenc K. MatConvNet: convolutional neural networks for MATLAB[C]//Proceedings of the 23rd ACM International Conference on Multimedia, October 26-30, 2015, Brisbane, Australia. New York: ACM, 2015: 689-692.
[19] Yu P, Qin A K, Clausi D A. Unsupervised polarimetric SAR image segmentation and classification using region growing with edge penalty[J]. IEEE Transactions on Geoscience and Remote Sensing, 2011, 50(4): 1302-1317.
[20] Liu B, Hu H, Wang H Y, et al. Superpixel-based classification with an adaptive number of classes for polarimetric SAR images[J]. IEEE Transactions on Geoscience and Remote Sensing, 2013, 51(2): 907-924.