Hyperspectral Image Zeroblock Coding Algorithm Based on 3D KLT and Wavelet Transform
Ying Min Hou, Long Liu, Lei Yang, Kang Xiao-long · 2010
Based on the characteristics of hyperspectral images, in this paper three-dimensional Set Partitioned Embedded Zero Block Coding (3D SPEZBC) algorithm based on spectral Karhunen-Loeve transform (KLT) and spatial wavelet transform (WT) for hyperspectral image compression is presented. This algorithm adopts ID KLT as spectral decorrelator and 2D WT as spatial decorrelator, the partitioning coding method based on the set representation structure to process each 2D subbands, and the context-based adaptive arithmetic coding scheme to further improve the coding performance. Furthermore, the compression performances of the low-complexity KLT are compared and evaluated, and then the best choices of coding parameters are given. In comparison with several state-of-the-art embedded wavelet coding algorithms, experimental results indicate that our coder can achieve better compression performance.