- Title: Optimizing PnP-Algorithms for Limited Point Correspondences Using Spatial Constraints
- Authors: Peter Roch, Bijan Shahbaz Nejad, Marcus Handte and Pedro José Marrón
- Abstract: Pose Estimation is an important component of many real-world computer vision systems. Most existing pose estimation algorithms need a large number of point correspondences to accurately determine the pose of an object. Since the number of point correspondences depends on the object’s appearance, lighting and other external conditions, detecting many points may not be feasible. In many real-world applications, movement of objects is limited due to gravity. Hence, detecting objects with only three degrees of freedom is usually sufficient. This allows us to improve the accuracy of pose estimation by changing the underlying equation of the perspective-n-point problem to allow only three variables instead of six. By using the improved equations, our algorithm is more robust against detection errors with limited point correspondences. In this paper, we specify two scenarios where such constraints apply. The first one is about parking a vehicle on a specific spot, while the second scenario describes a camera observing objects from a bird’s-eye view. In both scenarios, objects can only move in the ground plane and rotate around the vertical axis. Experiments with synthetic data and real-world photographs have shown that our algorithm outperforms state-of-the-art pose estimation algorithms. Depending on the scenario, our algorithm usually achieves 50% better accuracy, while being equally fast.
- Title: Visual Foreign Object Detection for Wireless Charging of Electric Vehicles
- Authors: Bijan Shahbaz Nejad, Peter Roch, Marcus Handte and Pedro José Marrón
- Abstract: Wireless charging of electric vehicles can be achieved by installing a transmitter coil into the ground and a receiver coil at the underbody of a vehicle. In order to charge efficiently, accurate alignment of the charging components must be accomplished, which can be achieved with a camera-based positioning system. Due to an air gap between both charging components, foreign objects can interfere with the charging process and pose potential hazards to the environment. Various foreign object detection systems have been developed with the motivation to increase the safety of wireless charging. In this paper, we propose an object-type independent foreign object detection technique which utilizes the existing camera of an embedded positioning system. To evaluate our approach, we conduct two experiments by analyzing images from a dataset of a wireless charging surface and from a publicly available dataset depicting foreign objects in an airport environment. Our technique outperforms two background subtraction algorithms and reaches accuracy scores that are comparable to the accuracy achieved by a state-of-the-art neural network (~97%). While acknowledging the superior accuracy results of the neural network, we observe that our approach requires significantly less resources, which makes it more suitable for embedded devices. The dataset of the first experiment is published alongside this paper and consists of 3652 labeled images recorded by a positioning camera of an operating wireless charging station in an outdoor environment.