Towards Localizing Structural Elements: Merging Geometrical & Semantic Detection in RGB-D

#visualslam #slam #robotics #vslam #IROS2024 #situational_awareness

Given the advantages of point clouds generated by RGB-D cameras, this paper aims to employ combined #depth and #visual information for geometrical calculation and #semantic confirmation, respectively. The pipeline initiates with plane extraction using the depth data provided in the point clouds of a given frame and completes with semantic label augmentation through processing its visual data, resulting in rich information for building component localization.

🔗 Links:
Link to the pre-print: https://doi.org/10.48550/arXiv.2409.0...
Accepted in #IROS24 (2024 IEEE/RSJ International Conference on Intelligent Robots and Systems), 14th Workshop on Planning, Perception and Navigation for Intelligent Vehicles