Joint Groupwise Image Registration and Fusion Based on Bounded Generalized Gaussian Mixture Model

In this paper, a joint groupwise registration and fusion approach is proposed for multi-source images. We apply a bounded generalized Gaussian mixture model (BGGMM) to approximate the joint intensity of multi-source images and the relationship from a fused image to the source images. The problem of joint groupwise image registration and fusion (IRF) is formulated by a maximum likelihood (ML) and it is performed by an expectation maximization algorithm. Extensive computer simulations verify both registration and fusion performance of the proposed approach. Empirical findings confirm that the performance of the proposed approach is significantly better than those of other conventional approaches.