Drones need to detect and localize each other if they are to collaborate in multi-robot teams or swarms. In this paper, a method based on dense optical flow (OF) is developed that detects dynamic objects. This is achieved by comparing the flow vectors with the direction to the Focus of Expansion (FoE) in the image plane. A simulation in AirSim is developed to validate this approach and to create a data set for motion-based dynamic object detection. This simulation includes ground-truth FoE, depth, OF and IMU data. The results show that our method performs well if the OF vector’s magnitude is large enough and its angle is sufficiently different from those of static world points. We expect that the presented method will serve as a useful baseline for deep learning methods using dense optical flow as input.
|Title of host publication||Proceedings of the 12th International Micro Air Vehicle Conference|
|Number of pages||7|
|Publication status||Published - 2021|
|Event||12th International Micro Air Vehicle Conference - Puebla, Mexico|
Duration: 17 Nov 2021 → 19 Nov 2021
Conference number: 12
|Conference||12th International Micro Air Vehicle Conference|
|Abbreviated title||IMAV 2021|
|Period||17/11/21 → 19/11/21|
FingerprintDive into the research topics of 'Motion-based mav detection in gps-denied environments'. Together they form a unique fingerprint.
Best Conference Paper Award of the 12th International Micro Air Vehicle Conference 2021
Li, S. (Recipient), de Wagter, C. (Recipient) & de Croon, G.C.H.E. (Recipient), 19 Nov 2021
Prize: Prize (including medals and awards)File