عنوان مقاله [English]
نویسنده [English]چکیده [English]
Since, monocular vision sensor cannot accurately measure relative range between the target and itself, it is difficult for UAVs equipped with this kind of sensors to track a target. This is exacerbated by the tracking a maneuvering aerial target in a three-dimensional space. To solve this problem, the accuracy of dynamic models is very important for predicting the movement and maneuvering of the target. This accuracy can be achieved by complicating and bringing the dynamic equations closer to the real dynamics of the vehicle, or by using appropriate models of the target acceleration. In this paper, we compare the performance three different dynamic models in tracking a mass-point maneuvering target. To do this, we used the UKF filter along with the monocular camera measurement model These models differ in terms of the coordinate system and the target acceleration model. The time spent and Root mean square error (RMSE) are considered as comparison criteria between these models.