摘要
针对静止摄像机近距离拍摄目标的情形,提出一种人体目标检测方法,主要包含两个步骤:在运动分割步骤中,利用长程和短程两次背景更新获取精确的背景图像,同时利用基于颜色空间的阴影判定方法消除阴影的干扰;在目标检测步骤中,将肤色检测与人体目标的几何比例经验值相结合,对运动目标进行判断。
To deal with the situation that the human objects are captured by a still camera in a short distance,a human object detection method is proposed in this paper.It mainly includes two steps：in the motion segmentation step,the background image is precisely obtained by using the long-term and short-term background updating algorithms each,and the shadows are cleared up via the colour space based shadow discrimination approach.In the object detection step,this method utilizes the complexion detection and the empirical value of height/width ratio of human body together to discriminate the motion objects.
出处
《计算机应用与软件》
CSCD
2010年第6期111-114,共4页
Computer Applications and Software
基金
广东省科技计划项目(2006B60155)
关键词
人体目标检测
运动分割
肤色判别
背景更新
Human object detection Motion segmentation Complexion discrimination Background updating
