In the field of computer vision, image match is one of the most important research topics. It is used to match two or more images, for example, at different time, from different sensors or from different viewpoints. In this paper, we present a robust method for realtime image registration. First, the feature points of two images were obtained by using Adaptive Harris detector, and the matching points of the two images are determined by using global search. And then the maximum correlation principle is involved to adaptively determine the optimal affine transformation parameters. In the stage of real-time video image registration, we can determine the searching area for the related feature points in the current frame using the affine transformation matrix between the reference image and the previous frame, which enables the matching method effective. The experimental results showed that the presented method achieved 25fps, which is fast enough, and the registration errors are small enough demonstrating the high precision of the proposed registration method.