Du Zheng-Jun, Huang Shi-Sheng, Mu Tai-Jiang, Zhao Qunhe, Martin Ralph R, Xu Kun
IEEE Trans Vis Comput Graph. 2022 Apr;28(4):1745-1757. doi: 10.1109/TVCG.2020.3028218. Epub 2022 Feb 25.
Accurate camera pose estimation is essential and challenging for real world dynamic 3D reconstruction and augmented reality applications. In this article, we present a novel RGB-D SLAM approach for accurate camera pose tracking in dynamic environments. Previous methods detect dynamic components only across a short time-span of consecutive frames. Instead, we provide a more accurate dynamic 3D landmark detection method, followed by the use of long-term consistency via conditional random fields, which leverages long-term observations from multiple frames. Specifically, we first introduce an efficient initial camera pose estimation method based on distinguishing dynamic from static points using graph-cut RANSAC. These static/dynamic labels are used as priors for the unary potential in the conditional random fields, which further improves the accuracy of dynamic 3D landmark detection. Evaluation using the TUM and Bonn RGB-D dynamic datasets shows that our approach significantly outperforms state-of-the-art methods, providing much more accurate camera trajectory estimation in a variety of highly dynamic environments. We also show that dynamic 3D reconstruction can benefit from the camera poses estimated by our RGB-D SLAM approach.
准确的相机位姿估计对于现实世界中的动态三维重建和增强现实应用至关重要且具有挑战性。在本文中，我们提出了一种新颖的RGB-D同步定位与地图构建（SLAM）方法，用于在动态环境中进行精确的相机位姿跟踪。以往的方法仅在连续帧的短时间跨度内检测动态成分。相反，我们提供了一种更精确的动态三维地标检测方法，随后通过条件随机场利用长期一致性，该方法利用了来自多帧的长期观测。具体而言，我们首先引入一种基于使用图割随机抽样一致性（graph-cut RANSAC）区分动态点和静态点的高效初始相机位姿估计方法。这些静态/动态标签被用作条件随机场中一元势的先验，这进一步提高了动态三维地标检测的准确性。使用TUM和波恩RGB-D动态数据集进行的评估表明，我们的方法显著优于现有方法，在各种高度动态环境中提供了更精确的相机轨迹估计。我们还表明，动态三维重建可以从我们的RGB-D SLAM方法估计的相机位姿中受益。
