作　　者： ; ; ; ; ; ; ;

机构地区： 中国科学院

出　　处： 《计算机工程与设计》 2015年第11期2963-2967,共5页

摘　　要： 为克服现有视频监控系统覆盖面积有限、单目摄像头定位多义性以及实时识别跟踪性能差的缺点,提出一种充分借鉴生物眼自然特性、多组仿生眼系统协同监控的广域周界视频监控系统。通过自主研制系统机械构造、电路控制系统,采用双目主动视觉定位、边界设定等方法,建立整个监控系统和边界的三维结构关系。搭建多组仿生眼系统并进行户外测试实验,实验结果表明,该设计拥有监控覆盖面积广、智能化高、稳定性好的特性。 To solve the existing video surveillance system＇s shortcomings of limited coverage,polysemy of monocular camera positioning and poor performance of real-time identification tracking,a wide-area perimeter video monitoring system was proposed which fully drew lessons from the nature and used multiple sets of bionic eye for collaborative monitoring.The mechanical structure and circuit control system were self-developed,and the 3Dstructural relations between monitoring system and boundary were set up by adopting the methods of binocular active vision and boundary setting.Several systems were designed and tested.The results show the system is characterized by wide coverage,high intelligence and good stability.

关 键 词： 仿生眼 双目摄像头 广域周界 系统协同 视频监控

领　　域： [自动化与计算机技术] [自动化与计算机技术]