The conventional methods of image matching are based on gray and shape of image. These two features can't describe image data well especially in tracking the faint and small targets. It's difficult to track the target with highprecision and steady. In this paper, a new algorithm of image matching was proposed. It extracts characters including gray and shape from image data and gains an integrated character using the information fusion algorithm with simplicity and effect. Then calculates the position coordinate of target using integrated characteristics by correlation degree. Simulation shows that the method is feasible and effective.