论文部分内容阅读
此算法能够实时地检测出复杂背景中车辆,并为以后的计数系统做好基础工作。背景差法是一种从视频序列中实时分割出移动物体的典型方法。本文提出了一种新的用于车辆检测的背景模型建模方法。实验结果表明,该系统能够快速地适应环境的改变,达到了实时的要求。
This algorithm can detect vehicles in complex background in real time and make basic work for the future counting system. Background subtraction is a typical method of segmenting moving objects in real time from video sequences. In this paper, a new background modeling method for vehicle detection is proposed. Experimental results show that the system can quickly adapt to changes in the environment and achieve real-time requirements.