Fixing a blurred photograph : image deblurring

Abstract

Image blur is an unwanted product and can be formed due to many reasons such as movement during image capture, low intensity during camera exposure, atmospheric problems etc. As a result, the image is degraded and the captured scene is rendered unreadable. In order to extract useful information from a blurred image, deblurring has to be done. Recently, blur removal has become one of the most common field of image processing research and has sparked a lot of interest.

In this project, we presented a detailed study on image deblurring, type of blur and a comparative study of various image deblurring techniques. We tested several these techniques in terms of performance, blur type, Peak Signal to Noise Ratio (PSNR).

This project involved the integration of image deblurring techniques provided by MATLAB image processing toolbox into a MATLAB application with friendly Graphical User Interface (GUI). The MATLAB application is created to provide more intuitive experience by showcasing some blurring and deblurring effects applied on input images.