Wavelet Based Multi-Dimensional Image Coding Algorithms: 3D-SPECK

Hyperspectral imaging is a powerful technique and hasbeen used in a large number of applications. However,it generates massively large image data sets. Accessand transport of these data sets will stress existingprocessing, storage and transmission capabilities. Anew embedded, block-based, wavelet transform codingalgorithm of low complexity is developed forhyperspectral image compression: Three-DimensionalSet Partitioned Embedded bloCK (3D-SPECK). 3D-SPECKefficiently encodes 3D volumetric hyperspectral imagedata by exploiting the dependencies in alldimensions. It can generate either SNR scalable orresolution scalable embedded bitstreams. It can alsogenerate ROI retrievable bitstreams.This bookintroduces 3D-SPECK and describes its technicaldetails. It is demonstrated that 3D-SPECK hasexcellent performance on hyperspectral image compression.

Dr. Xioali Tang was born in China. She received her Ph.D. degreefrom the Rensselaer Polytechnic Institute. She completedpostdoctoral training at the Massachusetts GeneralHospital/Harvard Medical School, and the University of Californiaat San Diego. Currently she is a faculty member with Universityof North Carolina.