An Integrated System Considering Object Detection and CNN-Based Classification for Damage Rank Estimation of Aging Bridges
摘要
Conventional damage assessments for bridge inspections are relied on subjective visual judgment by engineers. So, there are inconsistencies in evaluation and challenges in ensuring inspection accuracy. In this paper, we propose a system for damage evaluation and damage rank estimation utilizing inspection image data to establish an efficient and sustainable maintenance framework for aging bridges. The proposed system integrates damage recognition, classification and quantification based on damage images and rank information. The simulation results confirm the effectiveness of the proposed system by showing that damage ranks can be estimated from images of different cracks lengths.