Tags:Blur, Image de-blurring, Image Denoising, Noise, PSNR, RMSE and Wiener filter
Abstract:
There are many factors that lead to photographic image deterioration such as motion blur and geometric distortion. Image blurs results from the movement of the camera during the time the photo is being taken or the movement of the object to be photographed. Geometric distortion results from the use of a large angle lens. In this paper, a method for image de-blurring and image de-noising is presented by using an effective linear approach which is the wiener filtering. Initially, two images of peppers and a cameraman were used as the original image, then blurred and four different forms of noise (Gaussian, Salt & Peppers, Speckle and Poisson) were applied to the original image to perform noisy blurring im- age. The image is then removed from blurring and noise by using the Wiener filter. Wiener filters are designed and analyzed in this paper by using m-file MATLAB program. The results show that the wie- ner filter produces superior results since all of the blur is roughly eliminated. Furthermore, the results show that the wiener filter after de-noising performs better image quality for blur images and blur images with Poisson noise than the Wiener Filter after de-noising for images with Gauss- ian noise, Speckle noise, and Salt & Pepper noise respectively. The image quality parameters, PSNR and RMSE, provide greater performance for low SNR. The Tables show that the PSNR values are increasing while the RMSE values are decreasing.
Image de-Blurring and de-Noising by Using a Wiener Filterfor Different Types of Noise