摘要
本文提出了一种基于时间差分运动检测的改进方法。首先计算时间上连续的两帧灰度图像的差分图像;然后,将得到的差分图像通过低通滤波和二值化处理;最后,利用图像序列中的多帧灰度图像计算得到一组二值图像,通过对二值图像进行像素的算术运算,提取出运动目标区域。实验结果表明:这种方法是有效的。
An improved method of motion detection based on temporal difference is proposed in this paper. First, the difference image is got by two successive gray images. Then, the difference image is filtered and translated into binary image. Third, a number of binary images are calculated by some gray images in the image sequences. Finally, motion object regions are extracted by arithmetic operations of pixels from these binary images. Experimental results show that this improved method is suitable and effective.
出处
《南昌航空工业学院学报》
CAS
2005年第3期15-19,共5页
Journal of Nanchang Institute of Aeronautical Technology(Natural Science Edition)
基金
江西省自然科学基金(No.0411049)
江西省测控中心开放基金(No.ZX200308-16)
关键词
运动检测
时间差分方法
图像序列
算术运算
Motion detection
Temporal difference method
image sequences
Arith merle operations