System and method for 3D localization of landmarks from real-world images

A system is configured to implement a method for three-dimensionally positioning at least one landmark in a series of images from the real world, the images being provided by a continuously moving camera; the images are used in combined with image information and GPS information from a camera, camera parameters are unknown and estimated in the self-calibration step; wherein in a subsequent step, landmark localization is accomplished using one of camera self-motion and depth estimation. The camera is a monocular camera. 一种系统，设置用以实施一种用于对来自现实世界的一系列图像中的至少一个地标进行三维定位方法，图像由连续运动的摄像机提供；其结合使用来自摄像机的图像信息和GPS信息，其中摄像机参数是未知的，且在自标定步骤中被估计；其中在后续步骤中，地标的定位是利用摄像机自运动和深度估计的两者之一来完成的。摄像机是单目摄像机。