摘 要：基于视频的车辆检测技术在智能交通系统中的应用潜力巨大,但由于存在硬软件上的技术难点,仍有很多问题亟待解决,例如阴影消除等。在基于视频的车辆检测技术中,运用背景差分法获取前景车辆目标,针对选取小阈值可能产生的内外阴影,采用基于粗糙集阴影边缘点分类的方法,较好地实现了阴影边缘的检测,从而消除阴影。该方法简单,易理解及操作,精确度高。
关键词：视频检测；背景更新；阴影；粗糙集
中图分类号：TP391文献标识码：B
文章编号：1004-373X(2009)03-192-03
Video-based Techniques for Vehicle Detection and Shadow Eliminating Method
DU Lili,WANG Guoyu
(College of Information Science&Engineering,Ocean University of China,Qingdao,266100,China)
Abstract：Video-based techniques for vehicle detection have an enormous potential development in intelligent transportation system.But there′re still lots of difficult points needed to be solved in hardware and software techniques,such as shadow eliminating.Firstly,background auto-update's arithmetic is presented on the basis of image-differencing in order to obtain foreground objects.Then a novel method to classify shadow edge points is proposed which is based on rough sets.This method is simple,easily understanding and accurate.
Keywords：video detection;background update;shadow;rough sets
0 引 言
基于视频的车辆检测技术是在传统电视监视系统基础上发展起来的,以车辆检测技术、摄像机和计算机图像处理技术为基础,来大范围地对车辆施行检测和识别的新兴技术。（剩余1963字）