Compression of remote sensing image based on Listless Zerotree Coding and DPCM

The data quantity of remote sensing image is very large. Furthermore. the lowest frequency subband contains the main energy of original image and reflects the coarse of original image after remote sensing image is transformed by wavelet, so it is very important to the reconstructed image. Therefore a hybrid image compression method based on Listless Zerotree Coding (LZC) and DPCM is presented, namely, the lowest frequency subband is compressed by DPCM and others are compressed by LZC. LZC is a kind of zerotree coding algorithm for hardware implementation, which is based on SPIHT and substitutes two significant bit maps for three lists in SPIHT algorithm. Thereby LZC significantly reduces the memory requirement and complexity during encoding and decoding procedure. But LZC doesn't recognize the significance of grandchild sets, so the PSNR values of LZC are lower than SPIHT's and the compression speed drops. It is improved by adding a significant bit map that recognizes the significance of grandchild sets. A comparison reveals that the PSNR results of the hybrid compression method are 2 dB higher than those of LZC, and the compression speed is also improved.