摘要
本文介绍了一种基于单目序列图像的车辆三维信息的获取方法。在获取交通视频的基础上,根据摄像机的线性成像模型,对摄像机参数进行标定,建立空间中点与图像上点的映射关系。利用空间中点到图像上点的投影关系,以及车身上点在路面上的垂直投影,计算车身上特征点的三维坐标,进而恢复车辆的空间位置和三维信息。以实际交通录像为仿真对象,证明了该算法的有效性。
This paper introduces a method of acquisition of 3D information of vehicles based on monocular sequence images.On the basis of traffic video, calibrate the camera parameters according to the linear imaging model of camera, and get mapping relationship between image pixels and world points. By utilizing the 3D perspective mapping form the scene to the image, along with a plumb line projection, we can calibrate the 3-dimensional coordinates of the feature points on the vehicle body. Then, we can get the world location and the 3-dimensional information of the target vehicle. Experiments based on actual traffic video show that the algorithm proposed is valid.
出处
《电子设计工程》
2015年第5期116-119,共4页
Electronic Design Engineering
基金
国家"863"课题(2012AA112312)
关键词
单目视觉
摄像机标定
目标提取
三维信息
monocular vision
camera calibration
target detection
3D information
