Perth-CBD High-Resolution LiDAR Map with GNSS and IMU Calibration

This dataset provides a fully reconstructed, globally consistent 3D map of the Perth Central Business District (CBD), generated using high-resolution LiDAR, GNSS, and IMU data. The map is offered in .ply format and covers approximately 4.2 km\u00b2 of dense urban environment. It is constructed from 144,000 LiDAR frames collected with a 128-channel Ouster OS1-128 sensor and synchronized with RTK-GNSS and high-rate MEMS-IMU measurements across 21 carefully designed urban loops. Unlike the 2021 Perth-WA Localization Dataset (DOI: 10.21227\/s2p2-2e66), which only provided raw LiDAR sequences and frame-level poses for localization tasks, this new dataset offers a fully aligned and loop-closed 3D city model with accurate GNSS-IMU calibration and global geo-referencing. To ensure temporal and spatial consistency, LiDAR, GNSS, and IMU data are synchronized using Dynamic Time Warping (DTW) and refined via Extended Kalman Filtering (EKF). The global map is built through NDT-based local registration, pose graph optimization, and GNSS-constrained global alignment. Quality assurance is performed using intersection-level metrics and 2D cartographic comparisons, achieving a 61.4% reduction in alignment error compared to previous baselines. This dataset is designed to support advanced research in urban SLAM, sensor fusion, GNSS-denied navigation, city-scale change detection, and digital twin construction in complex metropolitan environments.