Detection of Cracking and Spalling Damage in Buildings and Bridges.

This is part of an NSF project (the information is this link: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2036193&HistoricalAwards=false}. In this project, over 2,200 images are used to label cracks and spalling on buildings and bridges damaged during extreme events. The data format is MS-COCO when the boundaries of the damage are manually drawn with polygon lines. The training and validation datasets are uploaded for the research community. The reference GitHub link is here: https://github.com/Bai426/Damage-Detection-with-COCO-data-and-Mask-R-CNN. If you think the data is useful to your research, please cite the following publications:
[1] Bai Y., Zha B., Sezen H., Yilmaz A. (2023). Engineering deep learning methods on automatic detection of damage in infrastructure due to extreme events. Structural Health Monitoring. 2023;22(1):338-352. doi:10.1177/14759217221083649
[2] Bai Y., Sezen H., Yilmaz A. (2021). "End-to-end Deep Learning Methods for Automated Damage Detection in Extreme Events at Various Scales," 2020 25th International Conference on Pattern Recognition (ICPR), Milan, Italy, 2021, pp. 6640-6647, doi: 10.1109/ICPR48806.2021.9413041.
[3] Bai, Y., Sezen, H., Yilmaz, A. (2021). Detecting cracks and spalling automatically in extreme events by end-to-end deep learning frameworks. ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 2, 161-168.
[4] Bai Y. (2022). Deep learning with vision-based technologies for structural damage detection and health monitoring. PhD dissertation at Ohio State University.

本数据集隶属于美国国家科学基金会（National Science Foundation，NSF）资助的项目，项目详情可访问链接：https://www.nsf.gov/awardsearch/showAward?AWD_ID=2036193&HistoricalAwards=false。该项目使用逾2200张图像，对极端事件中受损的建筑与桥梁的裂缝及剥落病害开展标注工作。当损伤边界通过多边形线条手动勾勒时，数据集采用MS-COCO格式。本项目的训练集与验证集已面向科研社区公开上传，相关参考GitHub仓库链接为：https://github.com/Bai426/Damage-Detection-with-COCO-data-and-Mask-R-CNN。若您认为本数据集对您的研究有所助益，请引用以下文献：
[1] Bai Y., Zha B., Sezen H., Yilmaz A. (2023). 极端事件下基础设施自动损伤检测的深度学习方法构建. 《结构健康监测（Structural Health Monitoring）》, 2023, 22(1): 338-352. DOI: 10.1177/14759217221083649
[2] Bai Y., Sezen H., Yilmaz A. (2021). 多尺度极端事件下自动化损伤检测的端到端深度学习方法. // 2020年第25届国际模式识别会议（ICPR）, 意大利米兰, 2021, 第6640-6647页. DOI: 10.1109/ICPR48806.2021.9413041
[3] Bai Y., Sezen H., Yilmaz A. (2021). 基于端到端深度学习框架自动识别极端事件中的裂缝与剥落病害. 《国际摄影测量与遥感学会摄影测量、遥感与空间信息科学年鉴（ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences）》, 2卷, 第161-168页.
[4] Bai Y. (2022). 基于视觉技术的深度学习在结构损伤检测与健康监测中的应用. 俄亥俄州立大学博士学位论文.

搜集汇总

背景与挑战

背景概述

该数据集包含2200多张建筑物和桥梁受损图像，采用MS-COCO格式标注裂缝和剥落损伤，用于深度学习方法的自动损伤检测研究。数据集提供了训练和验证集，并附有相关GitHub链接和引用文献。