论文部分内容阅读
本论文提出了一个利用二维时空图象进行交通自动监测的新方法。摄像机架设在公路上方，通过两个细缝检测窗口──垂直于道路方向的车辆检测窗和平行于道路方向的速度检测窗，便形成用于交通自动检测的二维全景图（ＰＶＩ）和外极面图（ＥＰＩ）。本文讨论时空图象生成和校准的方法，车辆检测的信息融合方法和速度估计的时空轨迹法。实验表明，基于低成本的硬件，可实时得到车辆计数、分类和速度计量等基本交通参数。
This paper presents a new method of automatic traffic monitoring using two-dimensional space-time images. The camera is erected above the road and forms a two-dimensional panorama (PVI) for automatic traffic detection and out-of-plane detection through two slit detection windows, a vehicle detection window perpendicular to the road and a speed detection window parallel to the road Extreme view (EPI). This article discusses methods for generating and calibrating spatio-temporal images, information fusion methods for vehicle detection, and spatio-temporal trajectory methods for velocity estimation. Experiments show that based on low-cost hardware, real-time vehicle count, classification and speed measurement and other basic traffic parameters.