基于边缘检测的人群密度估计

当前我们的社会是一个高速发展的社会。经济不断发展的同时，硬件水平也在不断的提高,我们的生活也被许许多多的高科技产品所包围,这些产品的的确确带给我们很多益处.在这样一个大的背景下,视频监控系统的应用变得无处不在。我们利用计算机图像处理技术来操纵视频监控,提高视频监控系统的自动化程度，尽量减少人为工作，这将是视频监控系统发展的未来方向。在我们的各种公共场所中，流动的人群变得越来越频繁,当人群密度偏高时,现场很不容易控制,往往还会出现人群踩踏事件,所以为了有一个安全的公共生活环境,对公共场所中的人群进行有效的管理和控制，是我们当前需要解决的。由此，频监控中指定区域内的人群密度估计方法就应运而生。
采用相机拍照的方式获得图像数据之后，将图像进行处理使得能在计算机上随意操作，接着对图像提取特征，边缘检测，使得能区分图像中的人与其他参照物，最后将人数统计出并得出了结果。
通过分析图像分割的原理和边缘检测的一些基本概念和算法，特别是对图像中人的边缘鉴定与提取做了深入研究,然后详细的阐述了视频监控中指定区域内的人数密度估计方法.由于我是研究是基于边缘检测的,所以经过灰度变换，边缘检测和人数统计这样的三个步骤进行，结合openCV，将算法转化为了代码，最后还对系统进行了仿真实验。
[主题词]  人数统计；图像分割；边缘检测：目标提取；

Population density is estimated based on edge detection

 
[Abstract]  The current society is a rapidly developing society. The economy continues to develop, the hardware level is also rising，we are lived with a lot of high technogical products,those things really benefits us very much .Under the major background, the application  of video surveillance systems become ubiquitous.With the improvement in the level of modernization of people's lives, Computer image processing technology to improve the degree of automation of video surveillance systems, minimize manual operations, will be the future direction of development of video surveillance systems. Especially in various public places in the flow of the crowd is more frequent, because too much crowds may cause security.How effective management and control of the people in public places, is a key issue that need to be addressed. As a result, the frequency of monitoring crowd density estimation method came into being within the designated area.
After acquire image data through camera, process it in order to operat on computer easily, then extract feature and edge detection so that computer can distinguish the crowd and background, finally count the amount.
Through some basic concepts and algorithms in the analysis of image segmentation and edge detection, exspicially in the edge detection, the and then described in detail the method of density estimation in the number of video surveillance within the designated area, the gray-scale transformation, edge detection and statistics on the number of three-step, combine with OpenCV, change the algorithm into code, and finally system simulation.
[Key Words]  Demographics; Segmentation; Edge Detection；Object Extraction;

*若需了解更多与协助请咨询↓→[电脑QQ][手机QQ]【数据协助】