Analysis of interscale and intrascale dependencies between image wavelet coefficients

We develop an information-theoretic analysis of dependencies between image wavelet coefficients. The dependencies are measured using mutual information, which has a direct link with data compression and estimation performance. Mutual information can be computed analytically for the special case where the image is a stationary autoregressive (AR)-1 Gaussian process. We have also developed methods to compute mutual information experimentally from wavelet domain image data. Our mutual-information analysis provides a mechanism for model evaluation and comparison