A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Indoor mapping using low-cost MLS point clouds and architectural skeleton constraints
Abstract MLS (Mobile Laser Scanner) offers adequate mobility and the ability to obtain accurate environmental data. However, it still faces some challenges, such as the heavy reliance on high-resolution LiDAR and the poor reliability of the mapping process, which limit its effectiveness in engineering projects. Therefore, we propose a novel, automatic and efficient indoor mapping method that utilizes low-cost MLS point clouds and architectural skeleton constraints. First, we determine the initial localization using designed architectural skeleton feature patterns and descriptors. Second, we register adjacent scans with different registration rules due to different abundance of architectural skeletons. Third, we correct cumulative error for both loop and non-loop areas. Last, we stitch scans into the indoor point cloud map, and build the novel lightweight point cloud map. Experiments are carried out in three typical large-scale scenes inside buildings, and results show that our method can produce accurate indoor maps automatically and efficiently.
Highlights Automatic indoor mapping via low-cost MLS in large-scale complex indoor scenes. Usage of various available architectural reference data. Cumulative mapping error correction for both loop and non-loop areas. A novel indoor map format – lightweight point cloud map.
Indoor mapping using low-cost MLS point clouds and architectural skeleton constraints
Abstract MLS (Mobile Laser Scanner) offers adequate mobility and the ability to obtain accurate environmental data. However, it still faces some challenges, such as the heavy reliance on high-resolution LiDAR and the poor reliability of the mapping process, which limit its effectiveness in engineering projects. Therefore, we propose a novel, automatic and efficient indoor mapping method that utilizes low-cost MLS point clouds and architectural skeleton constraints. First, we determine the initial localization using designed architectural skeleton feature patterns and descriptors. Second, we register adjacent scans with different registration rules due to different abundance of architectural skeletons. Third, we correct cumulative error for both loop and non-loop areas. Last, we stitch scans into the indoor point cloud map, and build the novel lightweight point cloud map. Experiments are carried out in three typical large-scale scenes inside buildings, and results show that our method can produce accurate indoor maps automatically and efficiently.
Highlights Automatic indoor mapping via low-cost MLS in large-scale complex indoor scenes. Usage of various available architectural reference data. Cumulative mapping error correction for both loop and non-loop areas. A novel indoor map format – lightweight point cloud map.
Indoor mapping using low-cost MLS point clouds and architectural skeleton constraints
Luo, Junqi (author) / Ye, Qin (author) / Zhang, Shaoming (author) / Yang, Zexin (author)
2023-03-11
Article (Journal)
Electronic Resource
English