Atmospheric Turbulence, Target Tracking,

Moving objects tracking in atmosphere turbulence is an important research task, and has broad applications in both military and unmilitary field. Combined with the National Natural Science Foundation of China project——"Study on the stability of the line of sight (SLOS) in atmosphere turbulence for optical active tracking", this thesis investigates how to reduce the effect of atmosphere turbulence on the performance of objects tracking system from two aspects which are atmosphere turbulence effects in atmosphere stochastic channel and tracking algorithm respectively.Adaptive optics (AO) is used in objects tracking system which works in atmosphere environment after analyzing the effects resulting from atmospheric turbulence. Considering the properties of the bending of laser beam in the slant path, the novel amendatory model of correcting the scintillation with AO is presented. The mathematics model of correcting the angle-of-arrival fluctuations with AO is established and the corrected short exposure optical transfer function (OTF) with AO is given. With numerical simulation, the experiment demonstrates that turbulence can be corrected by AO effectively.AO can correct the effects of atmosphere turbulence greatly, but it can't remove all of the turbulence noise. At the same time, other noise will be taken by AO inevitably. Therefore, one objects tracking algorithm which is fit for this condition is needed, while template matching tracking algorithm has advantages in tracking objects whose SNR is low and background is complex, and its reliability is good, but its deadly defect is that it can't meet real-time nature of objects tracking system due to its large numeration.The traditional template matching algorithm is improved, and the new template matching based on image distinct block processing is presented in this paper. The new algorithm not only reduces numeration, enhances the real-time nature, but also reduces other noise. The tracked object may change in the course of tracking, for example, flexing of its size, circumgyrating of its angle and so on, as a result, the tracking will make a mistake if we use the fixed template or the image which locates in the best matching position of the current frame to match in the next frame. So, it is necessary to update template real time in the course of dynamic tracking for continuous frames. This will improve reliability of objects tracking. In view of this problem, an adaptive template image matching tracking algorithm is put forward in which template updating parameter depends on the matching tracking confidence factor of among the frames and inner the frame in this paper. The simulative result indicates that the new template updating method has strong adaptability.