A novel template update method for IR seeker

In general, the centroid and block matching tracking algorithms have been used for target tracking systems such as IR seekers. However, the former may induce large tracking error in complex background due to low contrast between targets and background. Moreover, the latter can be suffers from variation of target size when the observer approach to a target. In this paper, we propose a simple and effective template update method for the block matching algorithm. Our proposed method performs the block matching algorithm using the original target template and the zooming target template. If the correlation obtained by the zoomed target template is larger than that of the original template, we update the original template. Therefore, our proposed algorithm can robustly track targets although target size varies.