Effective denoising and classification of hyperspectral images using curvelet transform and singular spectrum analysis

Qiao, Tong and Ren, Jinchang and Wang, Zheng and Zabalza, Jaime and Sun, Meijun and Zhao, Huimin and Li, Shutao and Benediktsson, Jón Atli and Dai, Qingyun and Marshall, Stephen (2017) Effective denoising and classification of hyperspectral images using curvelet transform and singular spectrum analysis. IEEE Transactions on Geoscience and Remote Sensing, 55 (1). pp. 119-133. ISSN 0196-2892 (https://doi.org/10.1109/TGRS.2016.2598065)

[thumbnail of Qiao-etal-TGRS-2016-Denoising-and-classification-of-hyperspectral-images-using-curvelet-transform]
Text. Filename: Qiao_etal_TGRS_2016_Denoising_and_classification_of_hyperspectral_images_using_curvelet_transform.pdf
Accepted Author Manuscript

Download (2MB)| Preview


Hyperspectral imaging (HSI) classification has become a popular research topic in recent years, and effective feature extraction is an important step before the classification task. Traditionally, spectral feature extraction techniques are applied to the HSI data cube directly. This paper presents a novel algorithm for HSI feature extraction by exploiting the curvelet transformed domain via a relatively new spectral feature processing technique – singular spectrum analysis (SSA). Although the wavelet transform has been widely applied for HSI data analysis, the curvelet transform is employed in this paper since it is able to separate image geometric details and background noise effectively. Using the support vector machine (SVM) classifier, experimental results have shown that features extracted by SSA on curvelet coefficients have better performance in terms of classification accuracies over features extracted on wavelet coefficients. Since the proposed approach mainly relies on SSA for feature extraction on the spectral dimension, it actually belongs to the spectral feature extraction category. Therefore, the proposed method has also been compared with some state-of-the-art spectral feature extraction techniques to show its efficacy. In addition, it has been proven that the proposed method is able to remove the undesirable artefacts introduced during the data acquisition process as well. By adding an extra spatial post-processing step to the classified map achieved using the proposed approach, we have shown that the classification performance is comparable with several recent spectral-spatial classification methods.