Reinforced concrete structure crack detection dataset 10519

A dataset comprising a total of 10,519 images depicting cracked surfaces and an equal number of images depicting non-cracked surfaces (indicating the absence of cracks in the entire image) was established from a previous data collection. These images in the dataset were manually labeled by the authors as “crack” and “intact (non-crack).”

This dataset is freely available for academic and research purposes, including direct use in training, validation, benchmarking, and comparative studies of damage detection and segmentation algorithms.

If you use this dataset in any publication, report, or derivative work, please cite the following paper:

Wang, J., Wang, P., Qu, L., Pei, Z., & Ueda, T. (2024). Automatic detection of building surface cracks using UAV and deep learning‐combined approach. Structural Concrete, 25(4), 2302-2322.

本数据集源自既往数据采集工作，共包含10519张带裂缝表面图像，以及等量的无裂缝表面图像（即整张图像无任何裂缝）。数据集内所有图像均由作者团队人工标注为「裂缝（crack）」与「完好无裂缝（intact (non-crack)）」。

本数据集可免费开放用于学术与研究场景，包括直接应用于损伤检测与分割算法的训练、验证、基准测试及对比研究。

若您在任何出版物、报告或衍生成果中使用本数据集，请引用以下论文：
Wang, J., Wang, P., Qu, L., Pei, Z., & Ueda, T. (2024). 采用无人机（UAV）与深度学习（Deep Learning）结合方法自动检测建筑表面裂缝. 《Structural Concrete》（结构混凝土）, 25(4), 2302-2322.