The Research Of Multispectral Images Compression Algorithms Based On EBCOT And BMMV

As the special application,the multispectral images are always demanded to be compressed with lossless compression,but the multispectral images spatial correlation is weak,the texture is also complicated and abundant,result in the compression ratios is low at large at present. Along with the remote sensing technology developing quickly,kinds of new spaceflight remote sensors and sense organs get large and large quantity data,which brings overcharge to the image storage and transmission,so as to hard to deal with. So,the research of the compression algorithms of the multispectral images is much significance in the reality.According to the characteristic of the mutlitspectral images and the coefficient characteristic of the images used wavelet transform,this paper brings forward a new scheme compressing the multispectral images,which is based on the BMMV algorithm and the improving EBCOT algorithm(NEBCOT).First,the new algorithm adopts 3DWT to eliminate the spatial and the spectral redundancy of the mulispectral images at one time. Although the images can be reduce the redundancy at a certain extent after the wavelet transform,there exists much correlation at the statistic meaning. Then,we make use of the BMMV algorithm and the NEBCOT algorithm(applying 2D NEBCOT to code the single–bond,utilizing the 3D NEBCOT to code multi–band),combining the ROI coding algorithm to code entropy to decrease the statistic correlation.The outcome of the test indicates that the speed of coding and decoding of the new algorithm is fast,the image compression ratios and renew quality is high,and it supports the image gradually transmission,which attains to the anticipative purpose.