一种低信噪比下点目标检测新算法  

New point target detection algorithm in low SNR

作　　者：刘志刚[1] 卢焕章[1] 陈辉煌[1] 

机构地区：[1]国防科学技术大学ATR实验室,湖南长沙410073

出　　处：《系统工程与电子技术》2004年第11期1588-1591,共4页Systems Engineering and Electronics

摘　　要：针对红外探测系统对远距离目标进行实时临控的应用场合,提出了一种快速点目标检测算法。该算法将原始图像序列分割成多个小段,通过递推使用速度匹配滤波,能快速得到图像序列中目标轨迹的能量,从而提高低信噪比下点目标的检测能力。与其它检测算法相比,新算法具有检测概率高、算法复杂低等特点。对算法的原理和性能进行了详细地分析,通过仿真实验证明了该算法的有效性。To detect targets far from a detector in real time for IRST, a fast algorithm for point target detection is presented. In the algorithm, the long image sequence is divided into many short image sections, and the energy of target track is got quickly by using Velocity Matching Algorithm recursively, so the performance of target can be improved. Compared with other detection algorithms, the new algorithms's target detection probablity is higher, and computation is simpler. The new algorithm's theory and performance are analyzed in detail, and the algorithm's validity is prvoed by simulation experiment.

关 键 词：检测前跟踪 点目标检测 低信噪比 递推速度匹配法 

分 类 号：TP75[自动化与计算机技术—检测技术与自动化装置]