The invention provides an unmanned aerial vehicle target tracking method and system, an unmanned aerial vehicle holder and an unmanned aerial vehicle, and relates to the technical field of unmanned aerial vehicles, after a target image is acquired, two confidence threshold values are introduced, and the characteristics that the tracking frame rate of a tracker based on a correlation filtering algorithm and deformation thereof is high, but the target is easy to lose are considered; in combination with the characteristics that a deep learning-based tracker is low in tracking frame rate but not easy to lose a target, two trackers are matched, finally, a final target tracking result is determined based on a comparison result according to comparison of tracking result confidence coefficients of different trackers and two confidence coefficient threshold values, and on the premise of ensuring the tracking frame rate, the target tracking efficiency is improved. The method effectively solves the problems of target loss and wrong tracking caused by fast target movement, temporary shielding and the like, enhances the visual target tracking capability, and gives consideration to the frame rate and the accuracy.
本发明提出一种无人机目标跟踪方法、系统、无人机云台及无人机,涉及无人驾驶航空器技术领域,在获取了目标图像后,首先引入两个置信度阈值,考虑基于相关滤波算法及其变形的跟踪器的跟踪帧率高,但容易跟丢目标的特点,结合基于深度学习的跟踪器跟踪帧率低,但不易跟丢目标的特点,将两种跟踪器配合,最后根据不同跟踪器跟踪结果置信度与两个置信度阈值的比较,基于比较结果确定最终目标跟踪结果,在保证跟踪帧率的前提下,有效地解决由于目标运动较快、短暂被遮挡等情况出现导致的目标丢失、跟错的问题,增强了视觉目标跟踪能力,兼顾帧率及准确性。
Unmanned aerial vehicle target tracking method and system, unmanned aerial vehicle holder and unmanned aerial vehicle
一种无人机目标跟踪方法、系统、无人机云台及无人机
27.12.2022
Patent
Elektronische Ressource
Chinesisch
Stabilizer support, unmanned aerial vehicle holder and unmanned aerial vehicle
Europäisches Patentamt | 2025
|Miniature unmanned aerial vehicle holder and unmanned aerial vehicle thereof
Europäisches Patentamt | 2025
|Europäisches Patentamt | 2020
|Europäisches Patentamt | 2021
|