Title :
Image restoration using Gibbs priors: boundary modeling, treatment of blurring, and selection of hyperparameter
Author :
Johnson, Valen E. ; Wong, Wing H. ; Hu, Xiaoping ; Chen, Chin-Tu
Author_Institution :
Inst. of Stat. & Decision Sci., Duke Univ., Durham, NC, USA
fDate :
5/1/1991 12:00:00 AM
Abstract :
The authors propose a Bayesian model for the restoration of images based on counts of emitted photons. The model treats blurring within the context of an incomplete data problem and utilizes a Gibbs prior to model the spatial correlation of neighboring regions. The Gibbs prior includes line sites to account for boundaries between regions, and the line sites are assigned continuous values to permit efficient estimation using a method called iterative conditional averages. In addition, the effect of blurring in masking differences between images and the effects of misspecifying the amount of blurring are discussed
Keywords :
Bayes methods; correlation methods; iterative methods; picture processing; Bayesian model; Gibbs priors; blurring; boundary modeling; correlation; hyperparameter selection; image restoration; iterative conditional averages; masking; picture processing; Bayesian methods; Context modeling; Degradation; Image restoration; Inference algorithms; Iterative methods; Positron emission tomography; Radiology; Smoothing methods; Statistics;
Journal_Title :
Pattern Analysis and Machine Intelligence, IEEE Transactions on