摘要
本文阐述了一种帧差方式的自适应背景提取算法.该算法先用当前帧(IK)和参考背景图像进行差分,获得第一次的背景更新;然后通过用当前帧和其帧间距为N的前两帧图像(IK-N和IK-2N)分别差分,对第一次更新后的背景再进行修正后取代原来的参考背景.实验结果表明用该算法提取和更新背景的速度快,具有较好的鲁棒性和准确性.
This paper elaborates an extraction algorithm of self-adaptive background based on frame differential mode. The first background updates is obtained through comparing the frame difference of the current frame （IK） and the reference background image. Then the difference is compared respectively through the current frame and the first two frames with N as their frame pitch （IK-N and IK-2N）, further modifying updates of the first background and then replacing the previous reference background. The experimental results show that this algorithm can speed up the extracting and updating of the background, and it has better robustness and accuracy.
出处
《漳州师范学院学报（自然科学版）》
2012年第1期48-52,共5页
Journal of ZhangZhou Teachers College（Natural Science)
关键词
背景提取
自适应算法
运动目标检测
图像处理
background Extraction
self-adaptive algorithm
moving objects detection
image processing
