Hyperspectral Image Classification Using MiniVGGNet

Yıl 2021, , 295 - 303, 20.10.2021

Hüseyin Fırat , Murat Uçan , Davut Hanbay

https://doi.org/10.53070/bbd.989102

Öz

Hyperspectral image classification is widely used in the analysis of remote sensing images. Recently, deep learning has been seen as the most effective method for hyperspectral image classification. Especially, Convolutional neural networks (CNN) are getting more and more attention in this field. CNN provides automated approaches that can learn more abstract features of hyperspectral images from spectral, spatial or spectral-spatial fields. In this study, a 3D CNN based MiniVGGNet network is proposed to take full advantage of the relationships between hyperspectral features and to increase the classification accuracy. With 3D CNN, spectral-spatial features are extracted simultaneously. With MiniVGGNet, the number of trainable parameters is reduced and the training time is shortened. In addition, principal component analysis (PCA) is used as a preprocessing method to reduce the computational complexity caused by the high dimensionality of hyperspectral images. In order to test the performance of the proposed method, applications were performed on remote sensing datasets of Indian Pines, University of Pavia and Salinas. The results were compared with different deep learning-based methods. Better classification performance is obtained by using the proposed method for hyperspectral image classification.

Anahtar Kelimeler

Principal component analysis, MiniVGGNet, Hyperspectral image classification, 3D CNN.

MiniVGGNet Kullanılarak Hiperspektral Görüntü Sınıflandırma

Öz

Hiperspektral görüntü sınıflandırma uzaktan algılanan görüntülerin analizinde yaygın olarak kullanılmaktadır. Son zamanlarda, derin öğrenme hiperspektral görüntü sınıflandırmasında en etkili yöntem olarak görülmektedir. Özellikle evrişimsel sinir ağları (ESA) bu alanda giderek daha fazla ilgi görmektedir. ESA, spektral, uzamsal veya spektral-uzamsal alanlardan hiperspektral görüntülerin daha soyut özelliklerini öğrenebilen otomatik yaklaşımlar sağlamaktadır. Bu çalışma kapsamında, hiperspektral özellikler arasındaki ilişkilerden tam olarak yararlanmak ve sınıflandırma doğruluğunu arttırmak için 3B ESA tabanlı MiniVGGNet ağı önerilmektedir. 3B ESA ile spektral-uzamsal özellikler eş zamanlı olarak çıkarılmaktadır. MiniVGGNet ile de eğitilebilir parametre sayısı azaltılmakta ve eğitim süresi kısaltılmaktadır. Ayrıca, hiperspektral görüntülerin yüksek boyutluluğundan kaynaklanan hesaplama karmaşıklığını azaltmak için ön işleme yöntemi olarak temel bileşen analizi kullanılmaktadır. Önerilen yöntemin performansını test etmek için Indian Pines, Pavia Üniversitesi ve Salinas uzaktan algılama veri kümeleri üzerinde uygulamalar gerçekleştirilmiştir. Sonuçlar, farklı derin öğrenme tabanlı yöntemlerle karşılaştırılmıştır. Hiperspektral görüntü sınıflandırması için önerilen yöntem kullanılarak daha iyi sınıflandırma performansı elde edilmiştir.

Anahtar Kelimeler

Temel bileşen analizi, MiniVGGNet, Hiperspektral görüntü sınıflandırma, 3B ESA.

Kaynakça

• Ben Hamida, A., Benoit, A., Lambert, P., Ben Amar, C. (2018) 3-D deep learning approach for remote sensing image classification. IEEE Transactions on Geoscience and Remote Sensing 56: 4420–4434. https://doi.org/10.1109/TGRS.2018.2818945
• Chen, Y.; Jiang, H.; Li, C.; Jia, X.; Ghamisi, P. (2016) Deep feature extraction and classification of hyperspectral images based on convolutional neural networks. IEEE Transactions on Geoscience and Remote Sensing 54: 6232–6251. https://doi.org/10.1049/iet-ipr.2019.1282
• Chen, Y., Lin, Z., Zhao, X., Wang, G., Gu, Y., 2014. Deep learning-based classification of hyperspectral data. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 7: 2094–2107. https://doi.org/10.1109/JSTARS.2014.2329330
• Firat, H., Hanbay, D., 2021. 3B ESA Tabanlı ResNet50 Kullanılarak Hiperspektral Görüntülerin Sınıflandırılması Classification of Hyperspectral Images Using 3D CNN Based ResNet50. 2021 29th Signal Processing and Communications Applications Conference (SIU), Istanbul-Tukey, pp.6–9. https://doi.org/10.1109/SIU53274.2021.9477899
• Hanbay, K., 2020. Hyperspectral image classification using convolutional neural network and two-dimensional complex Gabor transform. Journal of the Faculty of Engineering and Architecture of Gazi University 35: 443–456. https://doi.org/10.17341/gazimmfd.479086
• Huang, L., Chen, Y. (2020) Dual-Path Siamese CNN for Hyperspectral Image Classification With Limited Training Samples. IEEE Geoscience Remote Sensing Letters 18: 1–5. https://doi.org/10.1109/lgrs.2020.2979604
• Jia, J., Wang, Y., Chen, J., Guo, R., Shu, R., Wang, J. (2020) Status and application of advanced airborne hyperspectral imaging technology: A review. Infrared Physics & Technology 104: 103115. https://doi.org/10.1016/j.infrared.2019.103115
• Li, S., Song, W., Fang, L., Chen, Y., Ghamisi, P., Benediktsson, J.A. (2019) Deep Learning for Hyperspectral Image Classification: An Overview. arXiv 57: 6690–6709.
• Liu, Y., Gao, L., Xiao, C., Qu, Y., Zheng, K., Marinoni, A. (2020) Hyperspectral image classification based on a shuffled group convolutional neural network with transfer learning. Remote Sensing 12: 1–18. https://doi.org/10.3390/rs12111780
• Makantasis, K., Karantzalos, K., Doulamis, A., Doulamis, N. (2015) Deep supervised learning for hyperspectral data classification through convolutional neural networks. IEEE International Geoscience and Remote Sensing Symposium, Milan-Italy, pp: 4959–4962. https://doi.org/10.1109/IGARSS.2015.7326945
• Mingyi He, Bo Li, H.C. (2017) Multi-scale 3D deep convolutional neural network for hyperspectral image classification. 2017 IEEE International Conference on Image Processing, Beijing-China, pp: 3904–3908.
• Mohan, A., Meenakshi Sundaram, V. (2020) V3O2: hybrid deep learning model for hyperspectral image classification using vanilla-3D and octave-2D convolution. Journal of Real-Time Image Processing. https://doi.org/10.1007/s11554-020-00966-z
• Mou, L., Ghamisi, P., Zhu, X.X. (2017) Deep recurrent neural networks for hyperspectral image classification. IEEE Transactions on Geoscience and Remote Sensing 55: 3639–3655. https://doi.org/10.1109/TGRS.2016.2636241
• 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. https://doi.org/10.1016/j.isprsjprs.2017.11.021
• Roy, S.K., Krishna, G., Dubey, S.R., Chaudhuri, B.B. (2019) HybridSN: Exploring 3D-2D CNN Feature Hierarchy for Hyperspectral Image Classification. arXiv 17: 277–281.
• Üzen, H., Türkoğlu, M., Hanbay, D. (2021) Derin U-Net Ağ Mimarileri Kullanarak Yüzey Hata Tespiti Surface Defect Detection Using Deep U-Net Network Architectures. 2021 29th Signal Processing and Communications Applications Conference (SIU), Istanbul-Tukey, https://doi.org/10.1109/SIU53274.2021.9477790
• Zhong, Z., Li, J., Luo, Z., Chapman, M. (2018) Spectral-Spatial Residual Network for Hyperspectral Image Classification: A 3-D Deep Learning Framework. IEEE Transactions on Geoscience and Remote Sensing 56: 847–858. https://doi.org/10.1109/TGRS.2017.2755542