摘要
针对帧差法易产生空洞以及背景减法不能检测出与背景灰度接近的目标的问题,提出了一种将对称差分法和背景减法相结合的运动目标检测算法。首先利用对称差分法和背景减法分别得到两种差分图像,并用OTSU分割法(大津法)得到合适的阈值将这两种差分图像二值化,然后将得到的两种二值化图像进行或运算,最后利用图像形态学滤波得到准确的运动目标。
Considering disadvantages of that frame differencing produces the empty hole easily and background subtraction can’t detect the object whose gray level is close to the background’s,a method of motion detection combining symmetrical differencing and background subtraction is proposed.First,two different difference images with symmetrical differencing and background subtraction are gained respectively and two binary images can be gained by the appropriate threshold which comes from the method of OTSU.Then,the precise moving object can be extracted by applying OR operator between the two binary images.Finally,the precise moving object will be gained by morphological filter.
出处
《计算机工程与应用》
CSCD
北大核心
2010年第3期145-147,共3页
Computer Engineering and Applications
关键词
对称差分法
背景减法
大津法
symmetrical differencing
background subtraction
OTSU＇s method
