高帧频大视场多目标实时检测方法研究

Real-Time Detection Investigations for Multiple Targets in High Frame Rate and Large Field

针对高帧频大视场电视跟踪系统中多目标检测的实际需要研究了一种简单的目标检测方法──自适应邻域均值门限检测方法；通过分析，证明该方法能够在信噪比较低和背景干扰较强的情况下对全视场中多个弱小目标进行实时检测，并建立了该算法的实时处理系统；结果表明该方法是完全可行的。

The adaptive mean threshold method is investigated according to the practical requirements of target detection of the video tracking system. The method can be used to detect multiple weak targets in the whole video field in real time in low signal noise ratio and strong disturbance background. A real time processing system for the algorithm is set up. The results show that the method is practical.