Frame-to-frame restoration of diffusion images

Frame-to-frame image data are acquired in many applications such as Radar, Biomedical, and Television imaging. In many situations, the imaging phenomenon can be modeled by a diffusion process. Here, we consider the problem of recursive filtering of such images on a frame-to-frame basis. Also, a two-dimensional problem of restoration of blurred images can be solved by imbedding it in this three-dimensional reeursive filtering problem without blur. The model structure leads to a computationally feasible filtering algorithm achieving large reduction of dimensionality and is useful in real-time hardware simulation or generation of such blurred image data as might occur in a forward looking radar (FLR).