Image Blur Detection Using Local Power Spectrum

Abstract

In this paper work, blur detection of images is carried with local power spectrum. De blurring of image plays a important role in image processing and computer vision techniques. In deblurring of image, the first step is considered the input image as a motion blurred image. Our blur detection is based on block by block local mean calculation. After that find out the global mean for the blurred image, then comparison of local mean with global mean takes place. The experimental result shows that the robustness of proposed algorithm. The proposed method performing operations on image for detecting blurred regions. After that detected blurred content converted in to an un blurred region that shows the final output of this method.