Apparatus and method for estimating orientation of camera relative to ground

An iterative multi-image camera orientation estimation includes taking an image of a scene in front of a camera; detecting line segments in the scene; calculating a maximum likelihood (ML) camera orientation by rotating an X-Y-Z coordinate system of the camera to optimally align it with line segments in at least two of the frontal, lateral, and vertical orthogonal directions to maximize a likelihood target; estimating a maximum posteriori (MAP) camera orientation that maximizes a posteriori target so that the MAP camera orientation is an optimal value between the priori camera orientation and the ML camera orientation and is closer to the one with less uncertainty; iterating the multi-image camera azimuth estimation, and respectively setting the priori camera azimuth and the corresponding priori camera azimuth uncertainty as the calculated MAP camera azimuth and the corresponding uncertainty until the uncertainty is lower than a threshold value. 一种迭代式多图像相机方位估计，包括：拍摄相机前的场景的图像；检测场景中的线段；通过旋转相机的X-Y-Z坐标系以使其与正面、横向和