Detection and Tracking of a Moving Target for UAV Based on Machine Vision
Duo Qi, Zhe Li, Baoxiang Ren, Pengfei Lei, Xiaotian Yang · 2021
With the continuously development of the technology related to unmanned air vehicle (UAV), the automation and intelligent levels become higher and higher. In view of the actual needs of precise control and the specific requirements of accurate tracking competition, this paper designs and realizes a complete process of detection and tracking. First, the images of moving targets are collected by machine vision. Then, an improved Yolo V3 algorithm is applied to finish the target detection work and an achieved proportional integral derivative (PID) controller to track the moving target is designed. To verify the accuracy of the detection and tracking algorithm and the effectiveness of the complete process, a verification system is also built. Real flight test results show that the UAV can complete the task with a high dynamic target detection rate and accurate tracking rate.