摘要
研究分析目前主流的背景建模方法,并针对动态交通场景中车辆目标持续运动,背景出现的概率较大的特点,提出一种基于彩色视觉信息统计的背景建模算法。实验结果表明,该算法可以较好地提取背景,并有效区分前景和背景。
Researches and analyses the current mainstream background modeling methods. Considering that vehicles are always moving in the dynamic traffic scene, and the background pixel appears frequently, proposes a background modeling method based on visual color information statistics which can better extract the statistical background and classify the foreground from the image frame.
出处
《现代计算机》
2011年第23期22-26,41,共6页
Modern Computer
关键词
背景建模:彩色视觉信息统计:目标检测
Background Modeling
Visual Color Information Statistics
Object Detection
