Natural Image Prior Super-Resolution
Natural Image Prior-based Bayesian Image Super-Resolution
by Haichao Zhang et al.
The Proposed Natural image prior-based Bayesian Super-Resolution (NBSR) Framework.
Introduction
In this work, we present a Natural image prior based Bayesian image SR (NBSR) approach with a flexible high-order MRF model as the prior for natural images. The Minimum Mean Square Error (MMSE) criteria is used for estimating the HR image. A Markov Chain Monte Carlo (MCMC) based sampling algorithm is presented for obtaining the MMSE solution. The proposed method can not only enjoy the benefits offered by the flexible prior, but also has the advantage of making use of the probabilistic modeling to perform a posterior mean estimation, thus is less sensitive to the local minima problem as the MAP solution. Experimental results indicate that the proposed method can generate competitive or better results than state-of-the-art SR algorithms.
Advantages
1) the proposed method is a fully Bayesian approach, which can incorporate the a priori knowledge on the latent HR image as well as other uncertainties such as the noise level into the framework in a natural way;
2) a more advanced prior is used for capturing the HR image statistics rather than other very simple priors. Specifically, a high-order MRF model is used for modeling the natural image statistics, which, at the same time, enables generative sampling from it;
3) the Bayesian MMSE criteria is applied for estimating the HR image rather than the MAP criteria. The MMSE approach has several advantages over MAP. It does not require ad-hoc modifications as MAP to achieve desirable restoration performance. Also, it is less sensitive to the local minima in the solution space than MAP, especially when heavy-tailed priors are applied.
Results
1. Image SR Quality Comparison ( PSNR, zooming factor = 3 )
2. Color Image SR ( zooming factor = 3 )
NN
Shan et al.'s Fast SR
Yang et al.'s ScSR
Proposed NBSR
NN
Shan et al.'s Fast SR
Yang et al.'s ScSR
Proposed NBSR
NN
Shan et al.'s Fast SR
Yang et al.'s ScSR
Proposed NBSR
3. Color Image SR ( zooming factor = 4 )
NN
Shan et al.'s
Babacan et al.'s
Kim et al.'s
Yang et al's
Proposed
NN
Bicubic
Freeman et al.'s
Kim et al.'s
Fattal's
Glasner et al.'s
Shan et al's
Proposed
Related Publications:
Generative Bayesian Image Super-Resolution with Natural Image Prior
Haichao Zhang, Yanning Zhang, Haisen Li and Thomas S. Huang
IEEE Trans. on Image Processing IEEE TIP, 2012 Accepted.
Related Projects: