We examine the problem of finding the optimal weight of the fidelity term in variational denoising. Our aim is to maximize the signal to noise ratio (SNR) of the restored image. A theoretical analysis is carried out and several bounds are established on the performance of the optimal strategy and a widely used method, wherein the variance of the residual part equals the variance of the noise. A necessary condition is set to achieve maximal SNR. We provide a practical method for estimating this condition and show that the results are sufficiently accurate for a large class of images, including piecewise smooth and textured images.
|Number of pages||12|
|Journal||Lecture Notes in Computer Science|
|State||Published - 2005|
|Event||5th International Conference on Scale Space and PDE Methods in Computer Vision, Scale-Space 2005 - Hofgeismar, Germany|
Duration: 7 Apr 2005 → 9 Apr 2005