基于自适应谱段分组的超光谱图像压缩算法

Hyperspectral imagery compression algorithm based on adaptive band regrouping

针对庞大的超光谱数据与有限的卫星信道容量间存在的巨大矛盾，提出了一种新的基于自适应谱段分组的超光谱图像压缩算法．为了充分挖掘图像的谱间相关性，运用该算法对超光谱图像进行了预处理，通过谱段的自适应分组和预测参考帧的选取提高了压缩算法的编码性能，并结合谱间预测和位平面编码分别消除了超光谱图像的谱间和空间冗余．实验结果表明：与传统方法相比，在保证图像质量和较低计算复杂度的前提下，其压缩编码的平均峰值信噪比提高了约2．0～4．5dB．

There is a big contradictory between the limited communication capacity of satellite channel and large amount hyperspectral data. A novel lossy hyperspectral image compression scheme based on adaptive band regrouping is proposed. As to exploit spectrum correlation sufficiently, the proposed method pre-processes hyperspectral image by band regrouping and reference frame selection, and decorrelates the spectrum redundancy with inter-band prediction and compresses the prediction residuals with bit-plane coding. Experiments show that the proposed approach has a good performance in quality and complexity, and the average peak signal-to-noise ratio is increased about 2.0～4.5 dB.