An efficient feature fusion in HSI image classification
In recent times, the fusion of spatial relaxation with spectral data has achieved remarkable success in target classification methods. Spatial relaxation is a scheme which exploits the neighbourhood relationship between the pixels of an image to minimize the spatio-spectral distortion. Application of spatial relaxation with spectral data can lead to reduce the noise effect and increase the class characterization. Such methods can also be applied to estimate the posteriors of the probabilistic classifier, to increase the classifier’s final accuracy. In this paper, we have introduced an edge based feature fusion method which helps in characterizing the class labels of hyperspectral image (HSI) in a better sense. It is an iterative method which exploits the spatial information from an image in such a manner that it assumes the feature preservation in vertical and horizontal directions for each pixel. With combining subspace regression based probabilistic method, the proposed method gives better accuracy for benchmark HSI datasets. Before this, we have implemented a fast Bayesian subspace regression method to achieve the posterior probabilities, for our edge feature relaxation method. Finally, we have compared the results with some recently proposed methods, and \(\alpha \) expansion graph cut optimization method, which is an efficient technique to fuse the contextual knowledge in posterior probabilities.
KeywordsHigh dimension data sets Semi supervised learning Edge preservation function Anisotropic feature fusion Probability relaxation
We have shown the impact of EPS based method on different bands of real HSI. Graphcut (GC) optimization has performed with the help of C++ wrapper libraries provided by shai bagon. Indiana Pines, Salinas Valley, and Pavia University data is obtained from [Aviris-NASA (JPL); Gamba (Accessed: 2019-01-20)] while spectral signatures for the synthetic data set has obtained from USGS spectroscopy lab. Figure 15 shows the band-wise features obtained by using EPS method on HSI.
- Aviris-NASA(JPL) HD. Indian pines and Salinas Valley datasets. https://aviris.jpl.nasa.gov/. Accessed 20 Jan 2019.
- Borges, J. S., Marcal, A. R. S., & Bioucas-Dias, J. M. (2007). Evaluation of bayesian hyperspectral image segmentation with a discriminative class learning. In 2007 IEEE International Geoscience and Remote Sensing Symposium (pp. 3810–3813). https://doi.org/10.1109/IGARSS.2007.4423673.
- Cao, X., Zhou, F., Xu, L., Meng, D., Xu, Z., & Paisley, J. (2018). Hyperspectral image classification with markov random fields and a convolutional neural network. IEEE Transactions on Image Processing, 27(5), 2354–2367. https://doi.org/10.1109/TIP.2018.2799324.MathSciNetCrossRefzbMATHGoogle Scholar
- Chambolle, A. (2004). An algorithm for total variation minimization and applications. Journal of Mathematical Imaging and Vision, 20(1–2), 89–97. https://doi.org/10.1023/B:JMIV.0000011325.36760.1e.MathSciNetzbMATHGoogle Scholar
- Chen, J., Du, L., Li, J., Han, Y., & Gao, Z. (2017). Improved algorithm for hyperspectral data dimension determination. IOP Conference Series: Earth and Environmental Science, 57(1), 012044.Google Scholar
- Devika, S., & Chaitanya, S. M. K. (2016). Signal estimation of hyperspectral data using hysime algorithm. In 2016 International conference on research advances in integrated navigation systems (RAINS) (pp. 1–3). https://doi.org/10.1109/RAINS.2016.7764372.
- Du, Q., & Younan, N. H. (2008). On the performance improvement for linear discriminant analysis-based hyperspectral image classification. In 2008 IAPR workshop on pattern recognition in remote sensing (PRRS 2008) (pp. 1–4). https://doi.org/10.1109/PRRS.2008.4783168.
- Dubes, R. C., Jain, A. K., Nadabar, S. G., & Chen, C. C. (1990). Mrf model-based algorithms for image segmentation. In Proceedings of the 10th international conference on pattern recognition (vol i, pp. 808–814). https://doi.org/10.1109/ICPR.1990.118221.
- Gamba, P. P. Pavai university Rosis dataset. http://www.ehu.eus/ccwintco/index.php/Hyperspectral_Remote_Sensing_Scenes. Accessed 20 Jan 2019.
- Ghamisi, P., Benediktsson, J. A., & Sveinsson, J. R. (2014). Automatic spectral-spatial classification framework based on attribute profiles and supervised feature extraction. IEEE Transactions on Geoscience and Remote Sensing, 52(9), 5771–5782. https://doi.org/10.1109/TGRS.2013.2292544.CrossRefGoogle Scholar
- Gonzalez, C., Lopez, S., Mozos, D., & Sarmiento, R. (2015). Fpga implementation of the hysime algorithm for the determination of the number of endmembers in hyperspectral data. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 8(6), 2870–2883. https://doi.org/10.1109/JSTARS.2015.2425731.CrossRefGoogle Scholar
- Krishnapuram, B., Carin, L., Figueiredo, M. A. T., & Hartemink, A. J. (2005). Sparse multinomial logistic regression: Fast algorithms and generalization bounds. IEEE Transactions on Pattern Analysis and Machine Intelligence, 27(6), 957–968. https://doi.org/10.1109/TPAMI.2005.127.CrossRefGoogle Scholar
- Li, S. (2001). Markov random field modeling in image analysis. https://doi.org/10.1007/978-1-84800-279-1.
- Li, J., Khodadadzadeh, C., Plaza, A., & Jia, X. (2016). A discontinuity preserving relaxation scheme for spectral spatial hyperspectral image classi cation. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 9(2), 625–681. https://doi.org/10.1109/JSTARS.2015.2470129.CrossRefGoogle Scholar
- Song, W., Li, S., Kang, X., & Huang, K. (2016). Hyperspectral image classification based on KNN sparse representation. In 2016 IEEE international geoscience and remote sensing symposium (IGARSS) (pp. 2411–2414). https://doi.org/10.1109/IGARSS.2016.7729622.
- Sun, L., Wu, Z., Liu, J., & Wei, Z. (2013a). Supervised hyperspectral image classification using sparse logistic regression and spatial-tv regularization. In 2013 IEEE international geoscience and remote sensing symposium-IGARSS (pp. 1019–1022). https://doi.org/10.1109/IGARSS.2013.6721336
- Sun, L., Wu, Z., Liu, J., & Wei, Z. (2013b). Supervised hyperspectral image classification using sparse logistic regression and spatial-tv regularization. In 2013 IEEE international geoscience and remote sensing symposium-IGARSS (pp. 1019–1022). https://doi.org/10.1109/IGARSS.2013.6721336.
- Tian, Y., Guo, P., & Lyu, M. (2005). Comparative studies on feature extraction methods for multispectral remote sensing image classification. https://doi.org/10.1109/ICSMC.2005.1571322.
- Waske, B., van der Linden, S., Benediktsson, J. A., Rabe, A., & Hostert, P. (2010). Sensitivity of support vector machines to random feature selection in classification of hyperspectral data. IEEE Transactions on Geoscience and Remote Sensing, 48(7), 2880–2889. https://doi.org/10.1109/TGRS.2010.2041784.CrossRefGoogle Scholar
- Yıldırım, I., Ersoy, O. K., & Yazgan, B. (2005). Improvement of classification accuracy in remote sensing using morphological filter. Advances in Space Research, 36(5), 1003–1006. https://doi.org/10.1016/j.asr.2005.05.043. (atmospheric Remote Sensing: Earth’s Surface, Troposphere, Stratosphere and Mesosphere-I) .CrossRefGoogle Scholar
- Zhu, J., Li, K., & Hao, B. (2018). Image restoration by a mixed high-order total variation and regularization model. Hindawi Mathematical Problems in Engineering, 2018, 13.Google Scholar