基于候选轨迹统计模型的红外弱小点目标检测(英文)  

Detection Algorithm for Infrared Dim Point Target based on Statistic Model for Set of Trajectory

作　　者：李俊[1] 周凤岐[1] 周军[1] 卢晓东[1] 

机构地区：[1]西北工业大学精确制导与控制研究所,陕西西安710072

出　　处：《火力与指挥控制》2009年第4期80-84,共5页Fire Control & Command Control

基　　金：航天科技创新基金资助项目(06CASC0404)

摘　　要：建立候选目标轨迹集合,以候选目标轨迹集合为样本空间,利用沿候选轨迹的能量积分构造此空间上的概率分布,从而建立候选轨迹的概率统计模型。以此模型为基础,设计了一种弱小点目标检测方法。通过蒙特卡罗实验方法调节模型参数,使得算法的检测性能达到所提出的指标。该算法规避了以往算法中需要获得解析形式的候选轨迹密度函数的限制,因此能检测出沿任意方向作直线运动的点目标。最后,针对给定的弱小点目标观测图像序列,利用所提出的算法进行检测,仿真结果证明此算法是有效的。The trajectory＇s set of potential target is built, then the set is regard as a sample space, and based on it the statistic model is proposed using the energy integral along every trajectory to describe the probability distribution of the sample space. Based on the model, a new method of point target detection is proposed. Through Monte Carlo experiments the model parameter is adjust for the purpose that the algorithm＇s performances achieve the desired level. This algorithm dose not required the analytic fashion of density function of potential trajectories like former algorithms, so the target＇s trajectory along arbitrary direction can be detected by this algorithm. At last, the algorithm proposed is used to a series of observation images to detect the target in these images, and the simulation results show the method of point target detection is effect.

关 键 词：弱小点目标 统计模型 检测算法 蒙特卡罗实验 

分 类 号：TN911.73[电子电信—通信与信息系统]