A new approach for the restoration of noisy blurred images
H. Tang, L.W. Cahill · 1991
An effective practical approach for restoring an image that has been distorted by a linear system is presented. An adaptive average differential iterative filter is used to eliminate noise as well as preserve edges at a relatively low cost for real-time applications. A spatial recursive deblurring approach is required to reduce any blurring effects such as those caused by atmospheric turbulence, horizontal motion, and CCD (charge coupled device) interactions. Then noise filtering is used to further reduce noise and spurious oscillations. This approach can be used in a variety of problems, ranging from the most simple linear deconvolution of noiseless images to the problems of constrained, adaptive and nonlinear restoration, and is particularly suited for the restoration of noisy blurred images.>