Classification of Hyperspectral Images in Remote Sensing with Convolutional Neural Network

Year 2024, Volume: 5 Issue: 1, 28 - 40, 28.03.2024

Ali Gündüz Zeynep Orman

https://doi.org/10.48123/rsgis.1344194

Abstract

Machine learning and deep learning methods exhibit high performance in classifying hyperspectral images, enabling more accurate and efficient classification of the images.. In this study, an approach is proposed that utilizes a combination of 1-D and 2-D convolutional neural networks (CNN) technologies for the classification of hyperspectral images. In the proposed model, principal component analysis is used for data preprocessing, and the obtained data is divided into spatial and spectral parts. The proposed model utilizes a combination of 1-D and 2-D CNN technologies. The integration of two powerful network structures has provided the ability to manage the complexity of hyperspectral images and offer a more effective classification capability with lower resource consumption. After combining the outputs of the hybrid CNN layers, the classification success is increased with an attention mechanism. For the solution of the overfitting problem, in addition to the use of a series of dropout and normalization layers, the ideal learning rate value is determined as 0.001. The performance of the proposed model was tested on the Indian Pines, University of Pavia, and Salinas datasets, and the kappa accuracy values were measured as approximately 97%, 99%, and 99%, respectively. The suggested model was compared with well-known approaches introduced in recent years, and it demonstrated superior performance in terms of classification accuracy.

Keywords

Hyperspectral image, Hyperspectral image classification, Convolutional neural networks, Attention mechanism, Remote sensing

Evrişimli Sinir Ağı ile Uzaktan Algılamada Hiperspektral Görüntülerin Sınıflandırılması

Abstract

Makine öğrenmesi ve derin öğrenme yöntemleri, hiperspektral görüntülerin sınıflandırılmasında yüksek bir performans sergileyerek, görüntülerin daha hassas ve etkin bir şekilde sınıflandırılmasına olanak tanımaktadır. Bu çalışmada, hiperspektral görüntü sınıflandırması için 1-D ve 2-D evrişimli sinir ağları teknolojilerinin birleşimini kullanan bir yaklaşım önerilmektedir. Önerilen modelde veri ön işleme olarak temel bileşen analizi kullanılmıştır ve devamında elde edilen veri, mekansal ve spektral olmak üzere ikiye ayrılmıştır. İki güçlü ağ yapısının birleştirilmesi, hiperspektral görüntülerin karmaşıklığını yönetme ve daha etkili ve düşük kaynak tüketimli bir sınıflandırma yeteneği sunmuştur. Hibrit olarak kullanılan evrişimli sinir ağı katmanlarının çıktıları birleştirildikten sonra dikkat mekanizması kullanılarak modelin sınıflandırma başarısı arttırılmıştır. Aşırı öğrenme sorununun çözümü için bir dizi bırakma ve normalizasyon katmanları kullanımının yanı sıra ideal öğrenme oranı değeri 0,001 olarak belirlenmiştir. Önerilen modelin performansı, Indian Pines, Pavia Üniversitesi ve Salinas veri kümelerinde denenmiş ve kappa doğruluk değerleri sırasıyla yaklaşık olarak %97, %99, %99 olarak ölçülmüştür. Önerilen modelin sınıflandırma doğruluğunun, literatürde öne çıkan yöntemlerle elde edilen sonuçlara göre daha üstün olduğu gösterilmiştir.

Keywords

Hiperspektral görüntü, Hiperspektral görüntü sınıflandırması, Evrişimli sinir ağları, Dikkat mekanizması, Uzaktan algılama

References

• Ahmad, M., Shabbir, S., Roy, S. K., Hong, D., Wu, X., Yao, J., ... Chanussot, J. (2021). Hyperspectral image classification—Traditional to deep models: A survey for future prospects. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 15, 968–999.
• Alhichri, H., Alajlan, N., Bazi, Y., & Rabczuk, T. (2018). Multi-scale convolutional neural network for remote sensing scene classification. Proceedings of the 2018 IEEE International Conference on Electro/Information Technology, Rochester, MI, USA, 1–5. https://doi.org/10.1109/EIT.2018.8500107.
• Chen, X., Xiang, S., Liu, C.-L., & Pan, C.-H. (2014). Vehicle Detection in Satellite Images by Hybrid Deep Convolutional Neural Networks. IEEE Geoscience and Remote Sensing Letters, 11(10), 1797–1801.
• Chen, Y., Zhu, L., Ghamisi, P., Jia, X., Li, G., & Tang, L. (2017). Hyperspectral images classification with Gabor filtering and convolutional neural network. IEEE Geoscience and Remote Sensing Letters, 14(12), 2355–2359.
• Dong, H., Zhang, L., & Zou, B. (2019). Band attention convolutional networks for hyperspectral image classification. arXiv. https://doi.org/10.48550/arXiv.1906.04379
• Fang, B., Li, Y., Zhang, H., & Chan, J. C. W. (2019). Hyperspectral images classification based on dense convolutional networks with spectral-wise attention mechanism. Remote Sensing, 11(2), 159. https://doi.org/10.3390/rs11020159
• Gao, H., Yang, Y., Li, C., Zhou, H., & Qu, X. (2018). Joint alternate small convolution and feature reuse for hyperspectral image classification. ISPRS International Journal of Geo-Information, 7(9), 349. https://doi.org/10.3390/ijgi7090349
• Goetz, A. F., Vane, G., Solomon, J. E., & Rock, B. N. (1985). Imaging spectrometry for earth remote sensing. Science, 228(4704), 1147–1153.
• Hamida, A. B., Benoit, A., Lambert, P., & Amar, C. B. (2018). 3-d deep learning approach for remote sensing image classification. IEEE Transactions on Geoscience and Remote Sensing, 56(8), 4420–4434.
• Hang, R., Li, Z., Liu, Q., Ghamisi, P., & Bhattacharyya, S. S. (2020). Hyperspectral image classification with attention-aided CNNs. IEEE Transactions on Geoscience and Remote Sensing, 59(3), 2281–2293.
• Haut, J. M., Paoletti, M. E., Plaza, J., Plaza, A., & Li, J. (2019). Hyperspectral image classification using random occlusion data augmentation. IEEE Geoscience and Remote Sensing Letters, 16(11), 1751–1755.
• Hubel, D. H., & Wiesel, T. N. (1962). Receptive fields, binocular interaction and functional architecture in the cat’s visual cortex. The Journal of Physiology, 160(1), 106–154.
• Jiao, L., Liang, M., Chen, H., Yang, S., Liu, H., & Cao, X. (2017). Deep fully convolutional network-based spatial distribution prediction for hyperspectral image classification. IEEE Transactions on Geoscience and Remote Sensing, 55(10), 5585–5599.
• Landgrebe, D. (2002). Hyperspectral image data analysis. IEEE Signal Processing Magazine, 19(1), 17–28.
• Leng, J., Li, T., Bai, G., Dong, Q., & Dong, H. (2016). Cube-CNN-SVM: a novel hyperspectral image classification method. Proceedings of the 28th International Conference on Tools with Artificial Intelligence (ICTAI), San Jose, CA, USA, 1027–1034. https://doi.org/10.1109/ICTAI.2016.0158
• Li, J., Bioucas-Dias, J. M., & Plaza, A. (2011). Spectral–spatial hyperspectral image segmentation using subspace multinomial logistic regression and Markov random fields. IEEE Transactions on Geoscience and Remote Sensing, 50(3), 809–823.
• Li, J., Bioucas-Dias, J. M., & Plaza, A. (2012). Spectral–spatial classification of hyperspectral data using loopy belief propagation and active learning. IEEE Transactions on Geoscience and Remote Sensing, 51(2), 844–856.
• Li, J., Zhao, X., Li, Y., Du, Q., Xi, B., & Hu, J. (2018). Classification of hyperspectral imagery using a new fully convolutional neural network. IEEE Geoscience and Remote Sensing Letters, 15(2), 292–296.
• Lin, M., Chen, Q., & Yan, S. (2013). Network in network. arXiv. https://doi.org/10.48550/arXiv.1312.4400
• Makantasis, K., Karantzalos, K., Doulamis, A., & Doulamis, N. (2015). Deep supervised learning for hyperspectral data classification through convolutional neural networks. Proceedings of the IEEE International Geoscience and Remote Sensing Symposium (IGARSS), Milan, Italy, 4959–4962. https://doi.org/10.1109/IGARSS.2015.7326945
• Md Noor, S. S., Ren, J., Marshall, S., & Michael, K. (2017). Hyperspectral image enhancement and mixture deep-learning classification of corneal epithelium injuries. Sensors, 17(11), 2644. https://doi.org/10.3390/s17112644
• Melgani, F., & Bruzzone, L. (2004). Classification of hyperspectral remote sensing images with support vector machines. IEEE Transactions on Geoscience and Remote Sensing, 42(8), 1778–1790.
• Paoletti, M. E., Haut, J. M., Plaza, J., & Plaza, A. (2018). A new deep convolutional neural network for fast hyperspectral image classification. ISPRS Journal of Photogrammetry and Remote Sensing, 145, 120–147.
• Roy, S. K., Krishna, G., Dubey, S. R., & Chaudhuri, B. B. (2020). HybridSN: Exploring 3-D-2-D CNN feature hierarchy for hyperspectral image classification. IEEE Geoscience and Remote Sensing Letters, 17(2), 277–281.
• Tarabalka, Y., Fauvel, M., Chanussot, J., & Benediktsson, J. A. (2010). SVM-and MRF-based method for accurate classification of hyperspectral images. IEEE Geoscience and Remote Sensing Letters, 7(4), 736–740.
• Wang, Y., Song, T., Xie, Y., & Roy, S. K. (2021). A probabilistic neighbourhood pooling-based attention network for hyperspectral image classification. Remote Sensing Letters, 13(1), 65–75.
• Yu, S., Jia, S., & Xu, C. (2017). Convolutional neural networks for hyperspectral image classification. Neurocomputing, 219, 88–98.
• Zhao, W., Jiao, L., Ma, W., Zhao, J., Zhao, J., Liu, H., Cao, X., & Yang, S. (2017). Superpixel-based multiple local CNN for panchromatic and multispectral image classification. IEEE Transactions on Geoscience and Remote Sensing, 55(7), 4141–4156.