基于嵌入式平台的猫眼效应目标实时探测系统研究

Real-time detection system of cat-eye effect target based on embedded platform

在安保工作中,为快速发现狙击手,构建了利用狙击镜的猫眼效应原理来发现目标的系统。该系统不同于传统的采用连续激光扫描视场,而采用可调制半导体激光器为探测光源,在调制信号的控制下照射视场,CCD接收相应的主被动回波图像。并根据目标距离不同所成图像特征不同的特点,提出了分别针对远近目标的识别算法,最终在嵌入式DSP高速处理平台上进行图像处理实现。外场实验结果表明,主被动图像相结合,且利用远近目标不同的识别算法,有效提高了探测率,降低了虚警率,系统准确探测到了300m近目标和800m远目标,并达到10frames/s的实时处理速度。

In order to solve the problem of finding the sniper quickly in security,a system using the cat-eye effect principle of sniper scope to find a target is designed.The system is different from that of traditional design of lighting source.It is a system based on semiconductor modulated laser,which replaces CW laser source,and works with modulated signal to scan the field.A CCD is used to capture the active and passive echo image.According to the target echoes image at different distance,the algorithm can be implemented in embedded DSP high-speed processing platform for the near and far targets respectively.The results,obtained in the field experimental,showed that the new system increases the detection rate,and reduces the incidence of false alarms.The system accurately detected the target of 300 mand 800min distance with a real-time processing speed of 10 fps.