使用基于学习的视觉里程计的动态稠密 RGB-D SLAM
Dynamic Dense RGB-D SLAM using Learning-based Visual Odometry
AI总结：
提出基于TartanVO的稠密动态RGB-D SLAM流水线，利用光流分割动静像素仅融合静态点，并重渲染去除动态像素后迭代回传以优化位姿估计。
摘要：
我们提出了一种基于学习的视觉里程计 TartanVO 的稠密动态 RGB-D SLAM 流水线。TartanVO 与其他直接方法而非基于特征的方法类似，通过稠密光流估计相机位姿，这仅适用于静态场景并忽略了动态物体。由于颜色恒常性假设，光流无法区分动态和静态像素。因此，为了通过此类直接方法重建静态地图，我们的流水线利用光流输出来解决动态/静态分割问题，并仅将静态点融合到地图中。此外，我们重新渲染输入帧以移除动态像素，并迭代地将其传回视觉里程计以优化位姿估计。
英文摘要：
We propose a dense dynamic RGB-D SLAM pipeline based on a learning-based visual odometry, TartanVO. TartanVO, like other direct methods rather than feature-based, estimates camera pose through dense optical flow, which only applies to static scenes and disregards dynamic objects. Due to the color constancy assumption, optical flow is not able to differentiate between dynamic and static pixels. Therefore, to reconstruct a static map through such direct methods, our pipeline resolves dynamic/static segmentation by leveraging the optical flow output, and only fuse static points into the map. Moreover, we rerender the input frames such that the dynamic pixels are removed and iteratively pass them back into the visual odometry to refine the pose estimate.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
