Open AccessResolution restoration algorithm based on maximum a posteriori from Poisson-Markov distribution and blind multichannel deconvolution
Author(s) -
Chen Yi-Nan,
Weiqi Jin,
Lei Zhao,
Zhao Lin
Publication year - 2009
Publication title -
wuli xuebao
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.199
H-Index - 47
ISSN - 1000-3290
DOI - 10.7498/aps.58.264
Subject(s) - maximum a posteriori estimation , deconvolution , image restoration , algorithm , computer science , blind deconvolution , point spread function , poisson distribution , pixel , markov chain , a priori and a posteriori , stability (learning theory) , image resolution , artificial intelligence , image processing , image (mathematics) , mathematics , maximum likelihood , statistics , philosophy , epistemology , machine learning
Single input multichannel output (SIMO) imaging system which enhances the spatial resolution by resampling the same scene increasingly has large applications, and the corresponding image restoration algorithm for uncontrollable SIMO has became the subject of current research. By the assumptions of Poisson, Markov stochastic field and regulation of improved multichannel constraint, the multichannel blind image reconstruction is proposed based on maximum a posteriori rules. The algorithm doesn't need to know the prior knowledge such as character, type and distribution of the degraded point spread function (PSF) in each channel. The pixels and PSF are projected into the amplitude restrained set and energy invariable set, respectively, and their estimates converge to the global optimum solutions by the alterative iteration, and finally, the super-resolution image is restored. Algorithm simulations and the real micro-shift, micro-defocus experiments indicate that it has good restoration effect and stability at different signal noise ratios and for different PSF supports.