Point Cloud Dataset of Reinforced Concrete Bridges Captured with Matterport Pro3 Scanner in Japan

This dataset contains high-resolution point cloud data of nine reinforced concrete bridges located in rural areas of Japan, collected in December 2023 using the Matterport Pro3 terrestrial laser scanner. The scanner features a 360° horizontal field of view (FOV) and a 295° vertical FOV, operating with a 904nm wavelength laser beam. It achieves a measurement accuracy of <b>±20mm</b> at a distance of <b>10m</b> and captures up to 100,000 points per second.<br> The collected data was manually annotated, classifying the points into five classes, namely,<br>0: Abutment1: Girder2: Deck3: Parapet4: OthersThe proportions of points in each class across the whole dataset, ranked from largest to smallest, are “others”, “girder”, “deck”, “abutment”and “parapet”. Notably, the “others”class covers all remaining elements that do not belong to the above four categories, such as vegetation, utilities, drainage components, surrounding terrain, and other non-structural or auxiliary objects.Key characteristics of the dataset:Data Format: LASCoordinate System: Local, without georeferencingResolution: Coordinate scale value of 1 mmThis dataset was created to support research on automated dimension estimation of bridge components using semantic segmentation and geometric analysis. It can be utilized by researchers and practitioners in structural engineering, computer vision, and infrastructure management for tasks such as semantic segmentation, structural analysis, and digital twin development.