Onboard marker-less detection and localization of non-cooperating drones for their safe interception by an autonomous aerial system

Matouš Vrba, Daniel Heřt, Martin Saska - CTU in Prague


In this paper, a novel approach to 3D localization of flying objects a Micro Aerial Vehicle (MAV) is presented. The presented method utilizes a depth image from a stereo camera to facilitate onboard detection of drones, flying in proximity of nearby flying objects without using any kind of markers, which enables localization of non-cooperating drones. This approach strongly relaxes the requirements on the drones to be detected, and the detection algorithm is computationally undemanding enough to process images online onboard an MAV with limited computational resources, which allows using it in the control feedback of an autonomous aerial intercepting system (AAIS). Output of the detection algorithm is filtered using a 3D tracking algorithm utilizing multiple instances of a Kalman filter, which also serves to preserve temporal consistency of the detections and predict positions of the drones (e.g. to compensate camera and processing delays). We demonstrate the importance of the advances in flying object localization, presented in this paper, in an experiment with an intruder-interceptor scenario, which would be unfeasible using state-of-the-art detection and localization methods.

Cite as:

  • M Vrba, D Heřt and M Saska. Onboard Marker-Less Detection and Localization of Non-Cooperating Drones for Their Safe Interception by an Autonomous Aerial System. IEEE Robotics and Automation Letters 4(4):3402-3409, October 2019. PDF, DOI BibTeX

Video from a fully antonomous aerial interception of an intruding drone

Video from an experiment, comparing drone localization using two detection methods

Videos from previous work on the AAIS platform

Experiments on marker-less localization of MAVs using convolutional neural networks

Testing of the AAIS control and planning algorithms