摘要
在研究帧差法的基础上,给出了基于连续帧差的一种改进算法:将摄像机拍摄的彩色图像经中值滤波后,采用彩色图像差分模型在连续3帧之间作差分运算,初始分割出移动的飞机,然后设置阈值将图像二值化,经过形态学腐蚀、膨胀处理后,再对得到的两幅差分图像作相与运算来确定移动飞机的位置。实验表明,该方法可完成智能安全监控系统中移动飞机的检测和定位任务。
Multi-frame difference,an improved algorithm based on frame difference,is adopted to detect and track moving air craft. Firstly, the image captured by camera is median filtered. By using a color image difference model ,differential processing is performed between three continuous results. The moving airplane is initially segmented. And then, binary images are made by setting a threshold. After morphological erosion and dilation processing,an AND operation between the two results is performed to locate the moving aircraft. Experiments show that the intelligent safety monitoring system using this method can detect and locate moving aircraft.
出处
《电视技术》
北大核心
2007年第10期81-83,共3页
Video Engineering
关键词
运动目标检测
帧差法
智能监控
moving object detection
frame differentiation
intelligent monitoring
