Restoration techniques for image blur, pattern removal, and video

Abstract

The purpose of this project is to investigate three techniques in image restoration. These three methods are unique in that they utilize data from the "bad" image in the restoration process. The techniques will restore the following images: an image that is blurred beyond recognition, an image that has a pattern, creating a still image from video. The restoration for the blurred image will be by using a modified Weiner filter. The pattern removal technique uses a subroutine that isolate the pattern that then may be subtracted from the original image. The creating of a still image involves shifting the sub-frames.