矢量量化压缩算法在高光谱影像上的研究实现

针对高光谱影像光谱维的数据量大、传统影像压缩方法不易于保存光谱内信息的特点,对矢量量化数据压缩方法中码书设计和码字搜索两个关键技术进行详细地研究,提出针对高光谱影像压缩的改进方法,并在此基础上实现了对高光谱影像的矢量量化压缩算法。最后通过对不同波段组合的AVIRIS的高光谱数据的实验,从压缩后的压缩率、速率和失真率等方面进行观察和对比,证明矢量量化压缩算法对高光谱影像具有显著的压缩效果。

Research and implementation of vector quantization on hyperspectral image compression

Plenty of continuous spectral information in hyperspectral image can not easy to be availably kept by traditional com-pression method and this paper introduces vector quantization compression technology applied in hyperspectral data to hold the spectral information.Two critical technology including designing codebook and searching codeword are researched,and improved method especially applied in hyperspectral image is proposed.The vector quantization algorithm to hyperspectral image researched in the paper has been programmed and implemented.The result of the experiment based on AVIRIS date with different bands combined,which includes compression rate,speed and distortion rate,shows that the vector quantization algorithm has obvious im-pact to hyperspectral image data.