装载有视频采集设备的巡检列车已被应用于铁路环境监测中。提出了一种从前向运动的列车车载视频中生成铁路环境立体全景图的方法,该方法仅使用一个摄像设备即可获取整个场景的立体全景信息。基于场景的几何先验和标定的相机参数从视频序列中连续地抽取一组梯形窄带作为完全采样拼接区域,窄带的宽度根据车速自动调节,将所有窄带经单应性变换修正后拼接得到全景图像,并据此构建准确的几何模型来生成全景立体像对。实验结果表明,该方法能够快速有效地生成具有立体感的全景图。与原始的视频格式相比,生成的全景图所需存储空间小,具有更广阔的视野和更强的真实感,可实现铁路场景的快速访问和沉浸式浏览。 Patrol train loaded with video capture equipment has been used in railway environment monitoring. A method of generating railway environment stereo panorama from the forward moving train car video is proposed. The method can obtain the stereoscopic panorama information of the whole scene with only one camera. A set of trapezoidal narrowbands are continuously extracted from the video sequence based on scene-based geometric a priori and calibrated camera parameters. The width of the narrowband is automatically adjusted according to the vehicle speed, and all the narrowbands are corrected by homography to obtain a panoramic image , And build an accurate geometric model to generate a panoramic stereo pair. The experimental results show that this method can quickly and efficiently generate a panoramic view with a sense of perspective. Compared to the original video format, the generated panorama takes less storage space, has a wider field of view and is more realistic, enabling quick access to railroad scenes and immersive browsing.