Estimating camera pose is one of the main obstacles to implement Simultaneous Localization and Mapping (SLAM) technology. In this paper, we suggest a new phase-based SLAM technique for a single-camera single-projector Digital Fringe Projection (DFP) System. This method incorporates a second fixed global projector that projects fringe images and gray codes across a larger field of view. Utilizing conventional algorithms, the camera’s position relative to the global projector is accurately calculated, compensating for error accumulation issues in traditional Visual Odometry (VO) methods. This eliminates the need for additional hardware or computationally expensive self-correction algorithms. The proposed technique has the potential to replace VO as the primary method for camera pose estimation, allowing for seamless stitching of multiple images to create a precise 3D map.
Paper
Full text
Global 3D mapping with enhanced camera pose estimation using auxiliary projector
Semantic Scholar · Computer Science · 2024
Abstract
Estimating camera pose is one of the main obstacles to implement Simultaneous Localization and Mapping (SLAM) technology. In this paper, we suggest a new phase-based SLAM technique for a single-camera single-projector Digital Fringe Projection (DFP) System. This method incorporates a second fixed global projector that projects fringe images and gray codes across a larger field of view. Utilizing conventional algorithms, the camera’s position relative to the global projector is accurately calculated, compensating for error accumulation issues in traditional Visual Odometry (VO) methods. This eliminates the need for additional hardware or computationally expensive self-correction algorithms. The proposed technique has the potential to replace VO as the primary method for camera pose estimation, allowing for seamless stitching of multiple images to create a precise 3D map.