Publications by authors named "Zhan-Wu Ma"
Sensors (Basel)
September 2024
Article Synopsis
- Recent advancements in visual SLAM technology have improved accuracy, but current RGB-D systems struggle with measurement errors from depth cameras when estimating 3D positions of feature points.
- The paper proposes a method called CI-TEDM, which combines ORB features, triangulation uncertainty, and depth measurement uncertainty using a Covariance Intersection filter to enhance position estimation.
- An improved CI-TEDM method integrates wheel encoder data to address tracking issues in difficult environments, leading to a robust visual-inertial RGB-D SLAM system that improves localization accuracy in real-time applications.
