Multi-image blind deblurring using a coupled adaptive sparse prior

This paper presents a robust algorithm for estimating a single latent sharp image given multiple blurry and/or noisy observations. The underlying multi-image blind deconvolu-tion problem is solved by linking all of the observations to-gether via a Bayesian-inspired penalty function which cou-ples the unknown latent image, blur kernels, and noise lev-els together in a unique way. This coupled penalty function enjoys a number of desirable properties, including a mech-anism whereby the relative-concavity or shape is adapted as a function of the intrinsic quality of each blurry obser-vation. In this way, higher quality observations may auto-matically contribute more to the final estimate than heavi-ly degraded ones. The resulting algorithm, which requires no essential tuning parameters, can recover a high quali-ty image from a set of observations containing potentially both blurry and noisy examples, without knowing a prior-i the degradation type of each observation. Experimental results on both synthetic and real-world test images clearly demonstrate the efficacy of the proposed method. 1