A Novel Approach for Image Deblurring

Abstract

In the area of image processing blur removal is essential step in image quality enhancement .It also has real time applications, therefore it is necessary to have efficient method to remove blur. We have proposed a non linear blur model which simply models low light pixels. In this work we have applied Gaussian kernel instead of Laplacian kernel. The proposed method is developed in such a way that it automatically detects low light pixel from a given blurred image. It also suppress the ringing artifacts. The more accurate results are obtained on problematic and challenging blur images.