BridgeScan: LASER SCANNING FOR AUTOMATIC BRIDGE ASSESSMENT

The project “Laser Scanning for Automatic Bridge Assessment” called BridgeScan proposes the framework to use laser scanning to acquire 3D topographic data points of bridges and to process data for bridge assessment purposes. This project will involve (i) investigate a flight path and scanning parameters to scan a bridge structure by UAV with an integrated laser scanning sensor, (ii) develop automatic, robust, efficient methods to process LiDAR data to identify structural deficiencies for bridge condition rating and (iii) develop automatic methods for reconstructing a 3D model of the bridge for bridge assessment based on finite element analysis (FEA).

Goal of this WP is to investigate flight paths and associated flight parameters, and scanning parameters to capture the bridge with the best quality and quantity for damage detection and 3D geometric reconstruction.

This WP will convert raw 3D point cloud data into meaningful geometric information. A statistical model will be developed to select the best parameters and thresholds to improve the automatic level and accuracy of segmentation methods. Additionally, this WP aims also to extract point clouds of bridge components, for example, the bridge deck, piers and abutments.

This WP is to develop a new strategy for automatically identifying structural deficiencies and generating a 3D geometric model. The structural deficiencies can be expressed through structural deformation analysis, water bleeding, chemical attack, corrosion/crusting, spalling, scaling, and surface loss. Additionally, a point cloud of each bridge components identified in WP2. 2.2 will be used as input data to construct a 3D geometric model.

This WP will validate results of detected structural deficiencies changes in terms of omissions, commissions, correctness. The quality of the 3D models will be validated. Physical on-site inspectors will be deployed to inspect the bridge to compare to the bridge condition rating derived from a point cloud.