基于独立成分分析的高光谱图像有损压缩方法

Hyperspectral Images Lossy Compression Method Based on Independent Components Analysis

提出一种结合小波变换和独立成分分析(ICA)的高光谱图像有损压缩方法。采用最大似然估计与最大噪声分离相结合的方法对原始高光谱数据进行维数估计。依据维数估计的结果在光谱方向上采用ICA,在空间上运用离散小波变换。对于变换后的系数,使用多级树集合分裂算法和算术编码分别进行量化编码和熵编码。在机载可见光/红外成像光谱仪220波段高光谱数据上的实验结果表明,该算法可以在获得较高压缩率的同时,保留高光谱图像的光谱特性。

A lossy hyperspectral images compression algorithm based on Discrete Wavelet Transform（DWT） and Independent Component Analysis（ICA） is presented. Maximum Noise Fraction（MNF） method and Maximum Likelihood Estimation（MLE） are used to estimate dimensionality of original hyperspectral data. Based on the result of dimensionality estimation, ICA and DWT are respectively used in spectral and space direction. Set Partitioning In Hierarchical Trees（SPIHT） algorithm and arithmetic coding are respectively applied to the transformation coefficient, achieving quantify and entropy coding. Experimental results on Airborne Visible/Infrared Imaging Spectrometer（AVIRIS） 220 bands hyperspectral data show that the proposed method achieves higher compression ratio and more strong analysis capability than Comparative algorithms.