Multiple frame image restoration techniques that employ color constraints and incorporate image distortions

Multiple frame maximum a posteriori (MAP) image restoration methods are investigated that employ color constraints and that incorporate image distortions. Improvements in image resolution are demonstrated by applying well-established MAP image restoration techniques to multiple, randomly displaced microscopic images. To date, color MAP image restoration has not been attempted, but here new algorithms are developed to extend existing MAP techniques using cross correlations between any number of observed color images. Experimental verification is provided for three color bands. Modifications of the MAP algorithms are developed which incorporate distortions of the type experienced by wide field of view cameras. These modifications significantly improve restoration performance and extend the usefulness of the multiple image technique beyond current spatially invariant applications.