An iterative frequency-domain technique to reduce image degradation caused by lens defocus and linear motion blur

Imagery taken of the Space Shuttle is often degraded by lens defocus or image motion blur. Conventional space-domain filtering techniques allow enhancement of some features but also substantially increase noise. Frequency-domain techniques such as magnitude-phase merging or edge extraction often induce artifacts that degrade the image further. A procedure developed at the Video Digital Analysis System (VDAS) Laboratory at the Johnson Space Center approximates the modulation transfer function (MTF) of the input image from edge information within the scene and then utilizes an iterative technique to reduce the amount of blur. This work is intended to improve upon existing methods of image restoration given limited a priori scene information. These procedures were applied to a set of test images of the Space Shuttle obtained from film and video cameras at the launch site. Using these two sources gave a range of image quality over which to test the techniques. Separate procedures were developed to deal with linear motion blur and lens defocus. The iterative process allows the user to modify the frequency-domain filtering mechanism to achieve desired results. Output from these procedures showed a substantial visual improvement in image quality. Moreover, the techniques appeared to work as well on video as they did on film imagery. Future effort will be directed at reducing the ringing artifacts inherent in frequency-domain sharpening procedures