Image Quality Estimation for Blur Compensation Using Lucy - Richardson Method

Image Quality Estimation for Blur Compensation Using Lucy - Richardson Method

An important element of constructing the automated system of the image restoration is a measure of quality for restoration of blurred images in absence of any information related to the original image. Such systems operate only with distorted images and original images are unavailable. In the paper, the approach, based on the analysis of interline and intercolumn pixel correlation in the image restored by the Lucy - Richardson method, has been proposed. The studies have shown that using the proposed quality measure makes it possible to achieve good results of the blurred images restoration. The dominating problem of restoration - the estimate of distorting operator has been solved using the iteration approximation of the impulse response of the distorting system and the estimate of the approximation results using the proposed measure. As a result, the Lucy-Richardson method has been modified so that the restoration of images can be performed in the automated mode without a user intervention.
Kristina V. Breykina
JSC Research and Development Center «ELVEES», Moscow, Russia; National Research University of Electronic Technology, Moscow, Russia
Sergey V. Umnyashkin
National Research University of Electronic Technology, Moscow, Russia
Share