Automated Bridge Condition Assessment with Hybrid Sensing

Abstract

It is necessary to assess the physical and functional conditions of highway bridges at regular intervals to ensure they still meet their service requirements. Currently, this condition assessment is mainly performed through visual inspection, which has been identified with several limitations (e.g. the timeconsuming assessment process and heavy reliance on inspectors' personal experience). In order to overcome these limitations and enhance the current inspection practice, this paper presents a novel method for automated bridge condition assessment using a hybrid sensing system. Under the method, existing conditions of bridge components are first captured with a stream of point clouds and color images. Then, the bridge components and the defects inflicted on the components are detected utilizing their visual patterns. The detection results are mapped to the point cloud. This way, the 3D information of the components and defects can be retrieved. The bridge condition assessment can be made effectively through the 3D visualization of this information before carrying out any on-site detailed evaluations.