A Deformable Grid Approach for Bayesian Image Registration

M. Ceccarelli and M. Donatiello (Italy)


Markov Random Fields, Image Matching, Biomedical Im age Processing


The paper presents a novel non parametric image regis tration method based on an explicit representation of the warping function. The image registration problem is ap proached in the Bayesian framework with a prior term given by a Gaussian random field accounting the regularity of a deformable grid. The algorithm is completely unsu pervised, making it suitable for high throughput biomedical applications such as proteomics and cellular imaging.

Important Links:

Go Back