Robust multi-sensor image registration by enhancing statistical correlation
Kyoung Soo Kim, Jae‐Hak Lee, Jong Beom Ra · 2005
This paper deals with robust registration of the images acquired by different sensors, namely, the electro-optic (EO) and infrared (IR) ones. In this paper, we propose the two preprocessing schemes to improve the performance of normalized mutual information (NMI) based registration. Both schemes try to enhance the statistical correlation between a pair of EO and IR images for accurate and fast registration. The first scheme, extraction of statistically correlated regions (ESCR), extracts the regions in an image that are highly correlated to their corresponding regions in the other image. This extraction procedure is performed for each image, and the commonly extracted regions are used for calculating NMI. The second scheme, enhancement of statistical correlation by filtering (ESCF), adaptively filters out the pair of images to enhance the statistical correlation between them. The proposed schemes are applied to NMI-based registration and the results are prospective for various pairs of EO/IR sensor images in terms of registration accuracy, robustness, and speed.