AUTHOR=Shen Baohua , Jiang Juan , Li Daoguo , Qian Feng 

TITLE=Location and tracking of environmental pollution sources under multi-UAV vision based on target motion model

JOURNAL=Frontiers in Ecology and Evolution

VOLUME=Volume 11 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/ecology-and-evolution/articles/10.3389/fevo.2023.1178990

DOI=10.3389/fevo.2023.1178990

ISSN=2296-701X

ABSTRACT=In computer vision, the detection and tracking technology of moving objects has become a hot topic today. With the continuous maturity of UAV technology, quadrotor UAV is more and more widely used in the market, and its maneuverability and concealment are very strong. UAV visual recognition and mobile tracking technology uses UAV as a carrier, combined with machine vision technology, extracts and processes visual information, as the input of UAV autonomous intelligent control and the expansion of UAV terminal application field. The application of computer vision technology on drones is a new breakthrough, and has been widely used in traffic control, crop protection, drone tracking and shooting and other fields. The target tracking technology in this paper refers to the visual tracking of ground moving targets by the aircraft during flight. Since both the aircraft and the target are in motion, there are background and two motion vectors synthesized by the target and the background in the acquired picture. Therefore, this paper proposed a research on the location and tracking of environmental pollution sources under the multi-UAV vision based on the target motion model. The article firstly presented the UAV target tracking technology, and analyzed the development and types of UAVs in detail. Then, due to the increasingly serious pollution, this paper proposed to use UAV sensing technology to locate the pollution sources. Finally, in the experimental part, the UAV flight platform was tested, and the actual operation and positioning of the pollution source was carried out. The final experimental results showed that: within 0~360 seconds, the attitude angle obtained by the gradient descent method had no divergence phenomenon, which could effectively reduce the error caused by integration; the inclination angle deviation of the two groups of experimental equipment was within ±2.5°, the roll angle deviation was within ±3°, and the deflection angle was larger at certain moments, but the average deviation was only 0.8°. It also showed that the system could better adapt to the practical application requirements of quadrotor UAV.