公共场景监控下的人群密度估计是公共安全管理中的一个重要内容.针对大场景监控情况,研究了一种基于区域划分的智能化人群密度估计快速方法.在帧差提取出的彩色视频背景的基础上,按视觉比例将大场景分成子区域,然后对某个子区域采用像素统计的方法实现人群密度估计,其余子区域采用自适应比例的方法实现人群密度快速估计.实验结果表明该方法在实际应用中简单、有效、快速,能为公共场所的预警系统提供有力的帮助.%Crowd density estimation in public scene surveillance is an important issue of public security. A intellectualized method of fast crowd density estimation based on region division was studied to resolve the people number estimation of full big scene. The full big scene was divided into some needed sub-areas according to the visual scale after the color background had been extracted by the frame difference. The method of pixel statistics was used to estimate the density of a certain sub-area, and then a method of adaptive scale was used to realize the fast estimation of the other sub-areas. Experimental result shows that this method is simple, effective and fast in practical application. And it can also provide powerful help to the public warning system.
展开▼