Compression Of Multispectral Remote Image Based On Wavelet Transform

In the paper ,We discussed the characteristic of multispectral remote images obtained from Landsat-TM sensor.In the same time ,the spatial dependencies and spectral dependencies of the images and the characteristic of wavelet coefficient in the wavelet transform field are statistied and analysed.Some important conclusions are drawn by comparing multispectral image with ordinary image. In addition ,the wavelet transform and its applications in multispectral remote image are deeply discussed.The paper presents some new methods for the lossy compression of multispectral images,which combines the characteristic of multispectral images.One method based on subband quantity adopts combining the scalar and VQ to quantize the wavelet coefficient.The method realizes compression coding by combining SPIHT and Huffman coding.Another method based on classification prediction adopts lossy coding of the predictive coefficient by combining SPIHT and SPECK.The method adopts Huffman coding to do the lossless coding of codebook and index table .In the meanwhile it transmits all coding files to the decoding so as to reconstruct image .The quality of reconstructive image reaches the needs of remote sensing image application.Moreover,the classification method is used to compressing color image .The results of numerous experiments indict that combining classification and prediction is perspective in the compression of remote sensing image.In the end,the paper introduces the theory of unseparate wavelet ,and researchesunseparate wavelet with flex matrix of in the application of multispectral image.