With the development of Aerospace remote sensing technology, the hyperspectral remote sensing technology is widely applied in earth resources detection, environment investigation and military reconnaissance, etc. It is different from multispectral image, hyperspectral image spectrum period is as many as dozens to hundreds of band, with high resolution, large amount of data, corresponding to the data transmission problem, so the hyperspectral image compression is become necessary. In this paper the vector quantization algorithm is applied for hyperspectral image compression, first convert the three dimensional hyperspectral images into two dimensional pixel vectors, and then applied the vector quantization on the transformed pixel vectors. The experimental results show that the algorithm not only has better compression ratio, and can effectively save image spectrum section characteristics, which is a kind of efficient three-dimensional compression method.
Low-Complexity Principal Component Analysis for Hyperspectral Image Compression
