An approach for auto bridge inspection based on climbing robot

Abstract

Bridge inspection is a critical responsibility and essential task. Currently, many bridges should be inspected periodically for safety tests. Also, the maintenance and repair cost of bridges has rapidly increased every year, because the conventional methods such as bridge-detection vehicle and human visual inspection are time-consuming, expensive and lack quantitative objectivity. This paper proposes an approach and develops a system for auto bridge crack detection based on our well-developed bridge-climbing robot, which collects crack images with a high-resolution camera. In order to extract cracks and features automatically, a crack detection algorithm is proposed in this paper. There exits challenging issue on discriminating the real bridge crack from fake bridge crack, which does not harm the safety of the bridge. An approach of line structured light is proposed to address the issue. By combining the proposed crack inspection algorithm and line structured light method, the crack can be extracted and analyzed precisely. Some tests are conducted and both results verified the proposed inspection method and the developed system.