Regularized Simultaneous Super-Resolution with Automatic Determination of the Parameters

We derive a novel method for automatic determination of the regularization parameters applicable for the class of simultaneous super-resolution (SR) algorithms. The proposed method is based on the classical joint maximum a posteriori (JMAP) estimation technique, which is a fast alternative to estimate the parameters. Unfortunately, the classical JMAP technique can be unstable and generates multiple local minima. In order to stabilize the JMAP estimation, while achieving a cost function with a unique global solution, we derive an improved solution by modeling the JMAP hyper parameters with a gamma prior distribution. Experimental results illustrate the effectiveness of the proposed method for automatic determination of the regularization parameters for the simultaneous SR. We also contrast the proposed method to a reference method named KNOWN. KNOWN is a MAP based simultaneous SR algorithm where the parameters are fixed, either known a priori or extracted from the high-resolution frames which are not usually available in practice.