Restoring the Blur Image

Abstract

The Families, friends, professionals, and enthusiasts take countless numbers of photographs everyday, and inevitably, many images suffer from some sort of distortion, or blurring. In the proposed system of this paper, Fourier based image deblurring is implemented to recover the motion blurred images. For the blurred color images, the motion blur kernel is, first of all, estimated using motion length. The RGB color channels are partitioned into three channels (R, G and B) respectively. For each three channel, the Lucy-Richardson algorithm is separately applied as 2D image. The LR algorithm operate on iterative convolution and deconvolution processes to get the deblur version of the image. In order to speed up the system, the convolution and deconvolution processes are performed in frequency domain by means of Fourier transformation. The main fact of the LR algorithm is ringing artifacts around the strong edges of the image. The post processing stage of the system carries out the deringing by smoothing these ringing regions. Then, finally, the three separate R, G and B channels are superpositioned to get the resultant deblurred color image of the proposed system.