论文部分内容阅读
该文提出了一种适合于交通场景视觉监控实际应用中的方便的投影参数求取方法,主要着眼于方法的易用性,并在此基础上改进了模型可视化算法.它首先利用地平面与图像中的对应点建立一个对应矩阵（homogra-phy）,然后利用已知的摄像机高度以及地面上垂线在图像中的投影等约束信息来求取摄像机的参数.实验证明,该文提出的方法是有效的,而且能够比别的方法更方便地解决视觉交通监控实际中的摄像机标定和模型可视化问题.
In this paper, a convenient method for obtaining projection parameters is proposed, which is suitable for the practical application of traffic scene visual monitoring. The method mainly focuses on the usability of the method and, on the basis of this, improves the model visualization algorithm. Firstly, The corresponding points in the image are used to establish a homogra-phy, then the camera parameters are obtained by using the known camera height and the projection of the image on the vertical of the ground, etc. Experiments show that the proposed method Is effective and can more easily solve the problem of camera calibration and model visualization in visual traffic monitoring than other methods.