NOTE: Blind Restoration of Degraded Binary Markov Random Field Images

✍ Scribed by Bing Zhang; Mehdi N. Shirazi; Hideki Noda

• Publisher: Elsevier Science
• Year: 1996
• Tongue: English
• Weight: 415 KB
• Volume: 58
• Category: Article
• ISSN: 1077-3169
• DOI: 10.1006/gmip.1996.0007

✦ Synopsis

ori (MAP) estimate of the true image and use the iterated conditional modes (ICM) method [1] to generate a MAP The problem of restoring binary images degraded by regiondependent flip-flap noises is considered. The real image is deestimate. The same approach has been adopted in [2], scribed by a Markov random field (MRF). The iterative condiwhere the estimation is carried out iteratively by estimating tional modes (ICM) algorithm is adopted. It is shown that the parameters of the MRF model, used as a model for under certain conditions the ICM algorithm is insensitive to the true image, and the parameters of the region-depeneither the MRF model or the noise parameters. Using this dent flip-flap noises at each iteration. There are two drawproperty, we propose a blind restoration algorithm that has backs with this method. The first is that the estimation of the following two basic properties: (a) it does not require full the model and noise parameters is time-consuming; the knowledge about the model and noise parameters; (b) it can second is that the algorithm cannot be fully implemented be fully implemented in parallel. To show the effectiveness of in parallel. This is because in order to estimate the model the algorithm, simulation results are presented on the applications of the algorithm to degraded hand-drawn and synthetic and noise parameters, we need the whole intermediately images.