| 摘 要: 无人机在室内等复杂环境中飞行时，存在GPS信号较弱、惯性传感器累计误差较大等问题，导致无法实现室内的精准定位。本文提出一种基于粒子群圆检测算法的无人机目标定位方法，该方法通过OpenCV视觉模组进行图像预处理，并通过增量式PID(Proportion Integration Differentiation)与图像滤波相结合的粒子群圆检测算法获得目标标靶的核心坐标与半径等参数，通过对运动的实时控制、增加判断条件等提高无人机的标定效率与准确率。实验结果表明，该方法能够将无人机的位置调整时间控制在4 s以内，且将标定准确率提高到了90%，极大地缩短了无人机的标定时间，提升了无人机的位置标定准确率。 |
| 关键词: 粒子群圆检测；增量式PID算法；图像处理；目标定位；运动控制 |
|
中图分类号: TP272
文献标识码: A
|
| 基金项目: 2021 年上海市大学生创新创业训练计划项目(S202110252039). |
|
| UAV Target Positioning Algorithm based on Particle Swarm Circle Detection |
|
LI Wenqiang, WANG Jiadong, SUI Guorong
|
(College of Opto-electronic Information and Computer Engineering, University of Shanghai for Science and Technology, Shanghai 200093, China )
1935021414@st.usst.edu.cn; wangjd_0618@163.com; suigr@usst.edu.cn
|
| Abstract: When flying indoors or in other complicated environments, UAVs (Unmanned Aerial Vehicle) usually have problems such as weak GPS signals and large cumulative errors of inertial sensors, which make it impossible to achieve accurate indoor positioning. This paper proposes a UAV target positioning algorithm based on particle swarm circle detection algorithm, which preprocesses the image through OpenCV visual module, and obtains parameters such as the core coordinates and radius of the target through particle swarm circle detection algorithm combined with incremental PID (Proportion Integration Differentiation) and filtering.The calibration efficiency and accuracy of UAV are improved by realtime control of motion and the increase of judgment conditions. Experimental results show that the proposed algorithm can control the position adjustment time of the UAV within 4 seconds, and increase the calibration accuracy to 90%, which greatly shortens the UAV calibration time and improves the UAV position calibration accuracy. |
| Keywords: particle swarm circle detection; incremental PID algorithm; image processing; target positioning; motion control |
