光谱数据压缩。

简介：
Architecture for the Compression of Hyperspectral Imagery encompasses a range of techniques. Specifically, it includes lossless predictive compression methods applied to hyperspectral images, alongside lossless compression approaches utilizing linear prediction. Furthermore, research has explored lossless compression strategies for ultraspectral sounder data and locally optimal partitioned vector quantization of hyperspectral datasets. The pursuit of near-lossless compression of hyperspectral imagery has led to the development of adaptive DPCM methods incorporating crisp and fuzzy logic. Additionally, joint classification and compression strategies have been investigated for hyperspectral images, as well as predictive coding techniques specifically designed for this data type. Trellis-coded quantization offers another method for coding hyperspectral imagery, and three-dimensional wavelet-based compression is also employed. Spectral/spatial compression methods contribute to the broader field, and JPEG2000 is utilized for compressing Earth science data. Finally, addressing spectral ringing artifacts – a common issue in hyperspectral image data compression – remains an important area of study.