A New Method for the Discerning of Point Moving Targets´ Tracks

This paper is aimed to propose a new method for the discerning of point moving targets´ tracks to improve the current search strategies, which can not make full use of the entire trace information based on the domain of current point target location. By using the PRI search of least-square trace prediction, we establish a trace-fitting model based on probability and statistics. By setting the predicted position as the highest priority, the predication of movement direction as the criterion of prioritization, our method not only increases the search efficiency, but also reduces the drop alarm ratio. Then, we discuss the problem of the trace-fitting length limitation caused by the allowable continuous losing frames in the process of trace prediction. Through establishing the corresponding model, we improve the accuracy of the trace prediction, and obtain an optimal solution of the target trace-fitting number of frames.