Markov random field aided Bayesian approach for image reconstruction in confocal microscopy

The inverse problem associated with microscopic image reconstruction has attracted much attention and initiated a lot of interest in microscopic image processing. The inverse problem of image reconstruction from noisy and incomplete data is still one of the key factors for realizing the full potential of optical microscopy. We propose to address the inverse problem using Markov random field in the Bayesian domain. This approach has the potential advantage of incorporation of prior knowledge in the reconstruction process through the prior function, thus making the problem well-posed and computationally efficient for three-dimensional (3D) image reconstruction. Image reconstruction on 3D phantom and microscopic specimens shows comparatively noise-free and better-resolved images. The edges and minute features such as islets are well reconstructed. We believe that the proposed image reconstruction methodology will find applications in microscopy and imaging.