Journal
IEEE TRANSACTIONS ON VISUALIZATION AND COMPUTER GRAPHICS
Volume 28, Issue 4, Pages 1745-1757Publisher
IEEE COMPUTER SOC
DOI: 10.1109/TVCG.2020.3028218
Keywords
Cameras; Simultaneous localization and mapping; Three-dimensional displays; Visualization; Pose estimation; Dynamics; Robustness; RGB-D SLAM; dynamic SLAM; long-term consistency; conditional random fields; graph-cut RANSAC
Categories
Funding
- Natural Science Foundation of China [61863031, 61902210, 61521002]
- China Postdoctoral Science Foundation [2019M660646]
Ask authors/readers for more resources
Accurate camera pose tracking in dynamic environments is achieved in this article through a novel RGB-D SLAM approach, which utilizes long-term observations and conditional random fields for more accurate dynamic 3D landmark detection. Evaluation results demonstrate the superiority of this method over existing approaches.
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.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available