摘要
在常用的图像运动检测方法基础上提出了一种改进的运动目标检测算法,即将连续2帧差图像和背景差图像直接相乘,再将相乘的结果进行二值化处理得到运动检测结果,从而将运动目标从背景图像中分离出来,最终得到视频序列图像中运动存在与否的一个二值运动模板,提高了运动检测的效果。
Based on the normal video motion detection, this paper presents an improved detection arithmetic of motorial objects. First, we calculate the differences of the two successive frames and background, then acquire the multiplication of the two differences,and from these results we can separate the motion object from the background image and get a motion template which includes the motive information of the object. It improves the effect of motion detection.
出处
《河北科技大学学报》
CAS
2006年第1期89-92,共4页
Journal of Hebei University of Science and Technology
基金
河北省自然科学基金资助项目(F2005000077)
关键词
图像检测
运动检测
分块
帧差
image detection
motion detection
blocking
frame difference
