上升段飞行器目标的视频图像跟踪

Video Image Tracking of Ascending Segment Aircraft Targets

粒子滤波是一种基于贝叶斯估计理论和蒙特卡罗理论的实时目标跟踪方法,具有较为灵活的并行化跟踪方式,能够较好地维持跟踪目标的假设状态,具有较好的跟踪效果和鲁棒性。上升段飞行器目标飞行视频图像跟踪是火箭等目标飞行监控的重要阶段,但现阶段对飞行器上升段的视频图像跟踪主要依靠人工手动操作云台控制器,实现视频图像中的飞行器跟踪,跟踪图像存在跟踪滞后、画面抖动等现象,跟踪效果受人为因素影响较大。本文提出一种基于粒子滤波方法的上升段飞行器目标视频图像跟踪方法,建立飞行器目标粒子滤波跟踪模型实现对飞行器目标的识别和跟踪,在识别和跟踪的基础上建立云台控制模型,通过对云台的智能控制获得飞行器上升段的高质量图像。采用火箭发射的视频图像作为模型验证的实验数据,检验飞行器目标的跟踪效果。

Particle filter is a real-time target tracking method based on Bayesian estimation theory and Monte Carlo theory with flexible parallel tracking method,which can better maintain the hypothetical state of tracking targets,and has better tracking effect and robustness.The video image tracking of the ascending segment aircraft is an important stage of target flight monitoring such as rockets.However,the video image tracking of the ascending segment of the aircraft mainly relies on manual operation of the PTZ controller to realize the tracking of the aircraft in the video image,which leads to some phenomena such as tracking lag and picture jitter,and the tracking effect is greatly affected by human factors.In this paper,the video image tracking method of the ascending segment aircraft based on particle filter method is proposed.The target particle filter tracking model of the aircraft is established to realize the identification and tracking of the aircraft target.Based on the identification and tracking,the PTZ control model is established.Through the intelligent control of the gimbal,the high-quality image tracking of the rising section of the aircraft is obtained.Finally,the video image emitted by the rocket is used as experimental data for model verification to verify the tracking effect of the aircraft target.