Detection of Corrosion and Crack in the Structure of a Bridge Using Computer Vision Techniques

Abstract

Corrosion and cracks pose significant challenges to the structural integrity and safety of bridges. Early detection and monitoring of these defects are crucial to ensure timely maintenance and prevent catastrophic failures. In recent years, computer vision techniques have shown great promise in automating the inspection process. This paper proposes the utilization of two state-of-the-art object detection algorithms, YOLOv5 and Faster R-CNN, for corrosion and crack detection on bridges. A dataset comprising images specifically focused on cracks and corrosion is utilized, with bounding boxes meticulously annotating the regions of interest. Subsequently, the images in the dataset are employed to train both a Faster R- CNN model and a YOLOv5 model, enabling the detection of cracks and corrosion within the images. Comprehensive analysis is then conducted based on the outcomes of these detection models.