Catching objects in-flight is an outstanding challenge in robotics. In this paper, we present a closed-loop control system fusing data from two sensor modalities: an RGB-D camera and a radar. To develop and test our method, we start with an easy to identify object: a stuffed Piglet. We implement and compare two approaches to detect and track the object, and to predict the interception point. A baseline model uses colour filtering for locating the thrown object in the environment, while the interception point is predicted using a least squares regression over the physical ballistic trajectory equations. A deep learning based method uses artificial neural networks for both object detection and interception point prediction. We show that we are able to successfully catch Piglet in 80% of the cases with our deep learning approach.
Learning to Catch Piglets in Flight
2020
Fast Neural Perception and Learning for Intelligent Vehicles and Robotics workshop at IROS 2019
Preprint
Electronic Resource
English