Location and tracking of environmental pollution sources under multi-UAV vision based on target motion model
Xuebin Liu,Hanshan Li,Jingyun Xue,Tao Zeng,Xin Zhao
DOI: https://doi.org/10.1007/s00500-023-07981-9
IF: 3.732
2023-03-20
Soft Computing
Abstract:In computer vision, moving objects’ detection and tracking technology have become hot topics. With the continuous maturity of UAV technology, quadrotor UAV is increasingly widely used in the market, and their maneuverability and concealment are extreme. The application of computer vision technology on drones is a breakthrough and has been widely used in traffic control, crop protection, drone tracking and shooting, and other fields. Therefore, this paper proposed research on the location and tracking environmental pollution sources under the multi-UAV vision based on the target motion model. This article combines the target tracking technology of UAVs and the development and types of UAVs. Firstly, based on the diffusion model of pollutants in the river, a two-dimensional steady-state diffusion equation for pollutants in the river is established. Then, under the improved boundary conditions, the least squares method of the sum of squares of the measured data and theoretical values is used to model the target motion, and due to the increasingly serious pollution, this paper proposes to use UAV sensing technology to locate the pollution source. Finally, by combining pollution source location technology and drone tracking technology, the problem of not being able to quickly identify pollution sources leading to industrial waste discharge or accidental leakage during transportation poses a considerable threat to river safety. Finally, the use of pollution source positioning technology and drone tracking technology to solve the problem of failure to quickly identify pollution sources, resulting in industrial waste discharge or accidental leakage during transportation, poses a considerable threat to river safety. Experiments are carried out on the collected UAV data based on static and dynamic tests of UAV flight platforms, the experimental part, the UAV flight platform was tested, and the actual operation and positioning of the pollution source were carried out. The final experimental results showed that: within 0–360 s, 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 UAVs.
computer science, artificial intelligence, interdisciplinary applications