Multimodal and Multiband Image Registration using Mutual Information

In this paper, we present a novel histogram based method for estimating and maximising mutual information (MI) between two multi-modal and possibly multi-banded signals. Histogram based estimation methods are a common means for estimating the MI between two signals and the derivative of MI with respect to these signals. However, they do not scale well towards higher dimensions of the signals involved. We introduce a new estimation method which relies on the combination of non-uniform quantisation of the signal spaces and kernel density estimation to deal with this problem. Furthermore, we show how existing 1D-1D methods can be improved by using a combination of weighted histogram updates and kernel convolutions. These convolutions can be computed efficiently in the frequency domain, which reduces the computational cost significantly. The weighting scheme, on the other hand, enables us to compute analytical derivatives of MI with respect to either of both signals, which is important for further optimisation purposes. We illustrate our approach with several applications in parametric and non-parametric multi-modal image registration. Our case study is the registration of multi-band aerial images. More particularly, we demonstrate how optimisation of MI can successfully align intensity, infrared, natural color and pseudo-color images. However, the applicability of our algorithms is not confined to this particular domain. For example, several applications in the domain of medical imaging could potentially benefit from the proposed approach.