基于分类KLT的高光谱图像压缩

高光谱图像的有效压缩已经成为高光谱遥感领域研究的热点。提出了一种基于分类KLT（ Karhunen-Loeve Transform）的高光谱图像压缩算法。该算法利用光谱信息对高光谱图像进行地物分类，根据相邻波段的相关性对高光谱图像进行波段分组。在地物分类与波段分组的基础上，对每组的每一类地物数据分别进行KL变换，利用EBCOT（Embedded Block Coding with Optimal Trtmcation）算法对所有主成分进行联合编码。实验结果表明，该算法能够取得优于JPEG2000以及DWT-JPEG2000的压缩性能，适合实现高光谱图像的有效压缩。

Hyperspectral Images Compression Based on Classified KLT

Efficient compression of hyperspectral images has been the focus in the field of hyperspectral remote sensing. A new compression algorithm of hyperspectral images based on classified-Karhunen-Loeve Transform （KLT） is proposed. Ground classification of hyperspectral images is performed by using spectral information. Band grouping is carried out according to the correlation between adjacent two bands. Based on the ground classification and band grouping, KLT is performed on each ground class of hyperspectral images respectively in each group. EBCOT（ Embedded Block Coding with Optimal Truncation） algorithm is used for the joint coding of all the principle components. Experimental results show that the proposed algorithm can achieve better compression performance compared with those state-of-the-art compression algorithms such as JPEG2000 and DWT-JPEG2000 ,which is suitable for the efficient compression of hyperspectral images.