基于DSP的无人机编队视频跟踪技术

摘要

无人机编队飞行可协同完成任务,比单机执行任务效率更高.队形控制是实现编队飞行的关键技术之一.视觉跟踪技术能从视频序列中获得感兴趣目标的状态参数,故在编队中可以利用视觉信息进行目标检测与跟踪来确定邻机方位及距离信息,从而实现队形控制.首先利用视觉传感器获取无人机的运动图像,然后基于DSP采用KLT算法,计算出无人机在图像上的相对位移,实现无人机的跟踪.实验证明,跟踪结果准确性较高,满足实际应用的精度要求,可用于无人机编队稳定跟踪,为进一步解决无人机相对定位问题提供位移信息.%UAV formation flight task can be completed more efficiently than a single UAV to perform a task. The formation control is one of the key techniques in formation flight. The visual tracking technology can obtain the state parameters of the target from a video sequence. Therefore, the visual information can be applied to the formation to detect and track the target, so as to determine the orientation and distance information of the adjacent UAVs and realize the formation control. The moving images of UAVs is obtained by the aid of visual sensors, and then the KLT algorithm is adopted to calculate the relative displacement of UAVs on the images according to DSP, and realize the tracking of UAVs. Experiments show that the the technology has high tracking accuracy, can meet the accuracy requirements of practical application, can be used for UAV formation stable tracking, and can provide displacement information for overcoming the difficulty of UAV relative positioning.