Image denoising using shift-invariant techniques for camera images

This paper aims in presenting a thorough comparison of performance and usefulness of de-noising techniques, belonging to scale-space domain. Multi-scale Transform (MST) based image de-noising techniques overcome the limitation of Fourier based methods, as it provides the local and detailed information of non stationary image at different scales, which is necessary for de-noising process. The MST based image de-noising techniques, namely, Contourlet Transform (CT) and Non Subsampled Contourlet Transform (NSCT), have been selected for the de-noising of standard images. Further, the comparison of performance of each of the de-noising techniques have been carried out in terms of different noise variances and by using well known metrics, such as, Root Mean Square Error (RMSE) and Peak Signal-to-Noise Ratio (PSNR). Analysis of resultshows that shift-invariant NSCT technique outperforms the CT based de-noising technique in terms of both qualitative and quantitative evaluation.