摘要
针对复杂背景条件下运动小目标的检测 ,采用时域递归低通滤波和多帧相关的方法设计了一种运动小目标实时检测系统 .该系统适用于背景缓慢变化或背景处于相对静止的环境 ,能够在恒虚警概率条件下 ,很好地确定小目标的位置 .通过自适应阈值处理可以识别出小目标的轮廓 .
Temporal recursive low pass filtering algorithm and multi frame correlation were presented to detect samll moving targets in complicated background. A real time implementation hardware system was designed, which fits to background varying slowly or stationary with respect to the environment. Target and its region location were determined precisely with the condition of constant false alarm probability. Contours of small targers were identified with adaptive threshold.
出处
《红外与毫米波学报》
SCIE
EI
CAS
CSCD
北大核心
2000年第5期366-370,共5页
Journal of Infrared and Millimeter Waves
基金
国防科工委"八五"期间重大项目基金资助项目&&
关键词
实时检测系统
计算机视觉
运动小目标
temporal recursive low pass filtering, real time detection system, constant false alarm probability, adaptive threshold.
