A Theory for Multiresolution Signal Decomposition: The Wavelet Representation
IEEE Transactions on Pattern Analysis and Machine Intelligence
Constrained Restoration and the Recovery of Discontinuities
IEEE Transactions on Pattern Analysis and Machine Intelligence
Stochastic Relaxation, Gibbs Distributions, and the Bayesian Restoration of Images
IEEE Transactions on Pattern Analysis and Machine Intelligence
An adaptive reconstruction method involving discontinuities
ICASSP'93 Proceedings of the 1993 IEEE international conference on Acoustics, speech, and signal processing: image and multidimensional signal processing - Volume V
| Hi-index | 0.00 |
One of the basic problems in image restoration is to remove the additive noise generated by imaging systems while preserving sharp edges and details in the image. The goal of this paper is to present a restoration method based on a Markov random field (MRF) model with line-process in a multiresolution scheme. The Markov property of the intensity field takes into account local interactions between pixels, and the line process allows for the introduction of discontinuities in these interactions. Unfortunately, these models lead to nonconvex criteria. To reduce computation time, the deterministic graduated nonconvexity relaxation algorithm is used. his algorithm provides a good approximation of the global minimum. Multiresolution techniques yield a rapidly convergent algorithm and improve the visual quality of restored images. In order to adapt the model to the multiresolution scheme, a new local interaction function is used. The difficulty is to adjust parameters of the model across the resolutions. This method gives good results in terms of visual quality. Images are smoothed, while edges are preserved. We present some results on a very noisy image.