Statistical models for images: compression, restoration and synthesis

We present a parametric statistical model for visual images in the wavelet transform domain. We characterize the joint densities of coefficient magnitudes at adjacent spatial locations, adjacent orientations, and adjacent spatial scales. The model accounts for the statistics of a wide variety of visual images. As a demonstration of this, we used the model to design a progressive image encoder with state-of-the-art rate-distortion performance. We also show promising examples of image restoration and texture synthesis.