Interferential Multispectral Image Compression with Classified Weighted Rate-Distortion Optimization and Adaptive Coding Depth Control

A compression algorithm based on partial SPIHT with classified weighted rate-distortion optimization and adaptive coding depth control is proposed. The different significance weights are initially given to every coding passes in each zero tree, according to the category of interferential regions and the significance of coding bitplanes. Then we determine the coding depth of each zero tree by significance weights and depth control factor, and code zero trees independently by Partial SPIHT, which selects adaptively one of three coding modes relying on the probability of significant coefficients in each bitplane. Finally the classified weighted rate-distortion optimization truncation is applied within the coding depth, according to the distortion contribution to constructed spectrum of zero trees in different interferential regions. The experimental results show that the proposed algorithm improves the performance in spatial and spectral domain, and efficiently reduces the computation complexity and memory.