Energy efficiency in wireless sensor networks is a principal issue to prolong applications to track the movement of the large-scale phenomena. It is a selective wakeup approach that is an effective way to save energy in the networks. However, most previous studies with the selective wakeup scheme are concentrated on individual objects such as intruders and tanks, and thus cannot be applied for tracking continuous objects such as wild fire and poison gas. This is because the continuous object is pretty flexible and volatile due to its sensitiveness to surrounding circumferences so that movable area cannot be estimated by the just spatiotemporal mechanism. Therefore, we propose a cluster-based algorithm for applying the efficient and more accurate technique to the continuous object tracking in enough dense sensor networks. Proposed algorithm wakes up the sensors in unit cluster where target objects may be diffused or shrunken. Moreover, our scheme is asynchronous because it does not need to calculate the next area at the same time.