This paper presents the performance assessment of innovative model-based algorithms developed for pose estimation of uncooperative targets by processing sparse three-dimensional point clouds. This topic is of interest in the framework of advanced space applications, e.g., on-orbit servicing and active debris removal, which require a chaser spacecraft to execute autonomous relative navigation maneuvers in close-proximity of a space target. Both the problems of pose acquisition and tracking are addressed. The former one is carried out by combining the concepts of principal component analysis and template matching in order to limit both the computational effort and the amount of on-board data storage compared with traditional approaches. The latter is entrusted to a customized implementation of the iterative closest point algorithm, which adopts multiple model-measurement matching strategies as well as a refinement step, in order to respectively increase robustness and accelerate algorithm convergence. Also, safe transition from acquisition to tracking is implemented by means of autonomous detection of failures of the pose acquisition algorithm. The performance of the proposed techniques is investigated by means of numerical simulations in which the operation of an active LIDAR system as well as the target-chaser relative dynamics are realistically reproduced. Results demonstrate algorithms' effectiveness over a wide range of relative pose conditions and dealing with targets of variable size and shape, in spite of considerable sparseness of the measured datasets.

This paper presents a visual-based approach that allows an Unmanned Aerial Vehicle (UAV) to detect and track a cooperative flying vehicle autonomously using a monocular camera. The algorithms are based on template matching and morphological filtering, thus being able to operate within a wide range of relative distances (i.e., from a few meters up to several tens of meters), while ensuring robustness against variations of illumination conditions, target scale and background. Furthermore, the image processing chain takes full advantage of navigation hints (i.e., relative positioning and own-ship attitude estimates) to improve the computational efficiency and optimize the trade-off between correct detections, false alarms and missed detections. Clearly, the required exchange of information is enabled by the cooperative nature of the formation through a reliable inter-vehicle data-link. Performance assessment is carried out by exploiting flight data collected during an ad hoc experimental campaign. The proposed approach is a key building block of cooperative architectures designed to improve UAV navigation performance either under nominal GNSS coverage or in GNSS-challenging environments.

This paper presents a customized three-dimensional template matching technique for autonomous pose determination of uncooperative targets. This topic is relevant to advanced space applications, like active debris removal and on-orbit servicing. The proposed technique is model-based and produces estimates of the target pose without any prior pose information, by processing three-dimensional point clouds provided by a LIDAR. These estimates are then used to initialize a pose tracking algorithm. Peculiar features of the proposed approach are the use of a reduced number of templates and the idea of building the database of templates on-line, thus significantly reducing the amount of on-board stored data with respect to traditional techniques. An algorithm variant is also introduced aimed at further accelerating the pose acquisition time and reducing the computational cost. Technique performance is investigated within a realistic numerical simulation environment comprising a target model, LIDAR operation and various target-chaser relative dynamics scenarios, relevant to close-proximity flight operations. Specifically, the capability of the proposed techniques to provide a pose solution suitable to initialize the tracking algorithm is demonstrated, as well as their robustness against highly variable pose conditions determined by the relative dynamics. Finally, a criterion for autonomous failure detection of the presented techniques is presented.