Natural features enable tracking for Augmented Reality. But the conven-
tional method to measure features needs very high workload. Thus, in this study, we
have developed a method to measure natural features automatically for tracking. In the
developed method, the system gets images and shapes of the environment with a camera
and a LRF (Laser Rangender). The system changes the position with a robot and it
also calculate the location and the direction of the system using a registration method of
pointclouds with ICP (Iterative Closest Point). We have conducted two experiments to
evaluate the developed method.