Automatic optical and IR image fusion for plant water stress analysis

Automatic registration of an optical image and an associated IR image is a key step toward to automation of canopy temperature measurement in the process of plant water stress analysis. In this context, the scene of the IR image is completely included in the optical image and the transform between the two images may involve translation and rotation by a small angle, though a small scale difference may also be present. This automatic registration of data from two quite different imaging regimes presents several challenges, and is not susceptible to several common image processing techniques. In this paper, an automatic image registration algorithm, based on the fundamental cross correlation method is designed, which can avoid human intervention in the alignment process and is suitable for the application to plant water stress analysis where significant numbers of images need to be processed. The effectiveness of the software design is demonstrated via our experiments and the registration error performance is compared to the cases where the similarity criterion is replaced by that of mutual information and correlation ratio respectively.