The main objective of this research project was to investigate the applicability of mobile Light Detection and Ranging (LiDAR) and imaging sensors to help detect concrete cracks and displacement of railroad bridge components. This overall objective was decomposed into the following three research objectives:
1. Develop and evaluate prototype image processing algorithms for concrete crack detection and classification
2. Develop and evaluate three-dimensional (3D) models from LiDAR data to identify signs of bridge component displacements
3. Evaluate the effects to the image processing algorithms and 3D models with data collected using an unmanned aerial system (UAS) with integrated LiDAR and imaging sensors
To address the first research objective, prototype unsupervised image processing algorithms were developed to detect and classify longitudinal, transversal, and block cracks on concrete bridge surfaces (e.g., pile caps). The algorithms were evaluated using non-processed images collected with a UAS. The algorithms were 83% effective in correctly detecting and classifying concrete cracks.
To address the second and third research objectives, a mockup bridge structure was constructed using polyvinyl chloride (PVC) material to conduct experiments prior to conducting more expensive field tests. Experiments consisted of developing 3D models from collected LiDAR data at different controlled distances from the bridge structure, and evaluating the usability of the resulting models to help detect signs of bridge component displacements. The data acquisition approach employed for the second research objective was based on a controlled LiDAR sensor location using a tripod mount. For the third research objective, data were acquired using a customized UAS—completely built by the research team—with integrated mobile LiDAR sensor and data storage units, among other key components. Comparisons of resulting 3D models from the UAS data acquisition approach and those from the controlled sensor location approach indicated that there was no significant difference between them.
The results obtained from this research effort highlighted the potential practical value from using UAS and sensor technology for bridge inspection purposes. Potential payoffs for practice include improved safety and accuracy of inspections, and reduced inspection costs.
The final report is available.
