MODIS Image Compression with Optimal Inter-band Prediction and Integer Wavelet Transform

Abstract

In order to store and transmit MODIS (moderate resolution imaging spectro-radiometer) multispectral images effectively, a practical compression method using optimal inter-band prediction and integer wavelet transform (IWT) is proposed. First, MODIS intrinsic noises are removed. Then, an optimal inter-band predictor is designed for de-correlating the spectral redundancy, and an error image is produced. Next, spatial redundancy in the error image is removed by using IWT and SPIHT (set partitioning in hierarchical trees) algorithm. This method can implement lossless compression, quasi-lossless compression and lossy compression with high compression ratio. We demonstrate the compression performance with different compression schemes in details. The experimental results have shown that our proposed method is efficient and practical for MODIS images.