Blind restoration for defocus blurred image based on autocorrelation of derivative image

The research on the blind restoration for the images blurred by defocus effect has important meaning and actual metric. A super-resolution blind restoration algorithm for defocus blurred images was suggested according to the model of optical defocusing by using autocorrelation of derivative image. The autocorrelation of the second-order derivative blurred image received by Laplacian was computed, the radius of defocusing blur was be confirmed via the information contained in the result of the autocorrelation, and the restoration was finally achieved by MPMAP super-resolution method using the estimated defocusing radius as parameter. Experiments illuminate that the proposed algorithm can accurately estimate the radius of defocusing blur and restore clear images. Compared with other methods, the proposed algorithm declines the computation quantity, improves result precision and obtaines more detailed information by using the super-resolution restoration algorithm. It has been successfully applied into judging or appraisal of defocus images in practice works.