2020/09/08 by Kamil Żywanowski, Żywanowski, Kamil, Adam Banaszczyk +3 · 1 citation
Computer Science · Earth and Planetary Sciences · Engineering · Environmental Science · Physics and Astronomy · #3D Surveying and Cultural Heritage #Advanced Neural Network Applications #Advanced Optical Sensing Technologies #FOS: Computer and information sciences #Industrial Vision Systems and Defect Detection #Remote Sensing and LiDAR Applications #Robotics (cs.RO) #Robotics and Sensor-Based Localization
paper · pdf · doi:10.48550/arxiv.2009.03705
openalex publication_date 2020/09/08 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Loop closure based on camera images provides excellent results on\nbenchmarking datasets, but might struggle in real-world adverse weather\nconditions like direct sun, rain, fog, or just darkness at night. In automotive\napplications, the sensory setups include 3D LiDARs that provide information\ncomplementary to cameras. The presented article focuses on the evaluation of\ncamera-based, LiDAR-based, and joint camera-LiDAR-based loop closures applying\na similar processing pipeline consisting of a neural network under varying\nweather conditions using the newly available USyd dataset. The experiments\nperformed on the same trajectories in diverse weather conditions over 50 weeks\nprove that a 16-line 3D LiDAR can be used to supplement image-based loop\nclosure to increase loop closure performance. This proves that there is a need\nfor more research into loop closures performed with multi-sensory setups.\n