MODIS Image Compression with Optimal Waveband Prediction Tree

Proposed a practical compression method using optimal waveband prediction tree, integer wavelet and SPIHT algorithm for MODIS multispectral images effectively. First, an optimal waveband prediction tree is designed for de-correlating the spectral redundancy. Then, spatial redundancy in the image is reduced by using IWT and SPIHT algorithm. Finally the arithmetic algorithm is used to improve the compression performance. We demonstrate the compression performance with different compression schemes in details. The experimental results show that the method is efficient and practical for MODIS images.