Near-lossless Compression of Hyperspectral Image Based on Adaptive Prediction
一种基于自适应预测的高光谱图像近无损压缩方法

A near-lossless compression algorithm was presented.First,3D adaptive prediction was introduced in to remove spatial and spectral redundancies efficiently.Then prediction errors were quantized properly to reduce coding rate further.Experiments show that this method can reduce bit-rate distinctly while reconstruction error can be controlled.