2026-10-11T20:51:17Z https://data.mendeley.com/oai

oai:data.mendeley.com/r8zrzk5w8j.2 2026-09-15T09:04:40Z

Zamanzade Nasrabadi, Amir Mohammad Fresh Concrete Surface Crack Detection Dataset: 659 Field Images with Pixel and Instance Annotations Mendeley Data Version 2 Update: - Reorganized dataset structure with explicit train/valid/test splits - Added separate folders for YOLOv8 polygon annotations (labels_yolo/) and U-Net binary masks (masks_unet/) - Included data.yaml configuration file for direct use with Ultralytics YOLOv8 - Added comprehensive README.txt with dataset documentation - Total: 659 images (461 train / 132 validation / 66 test) with both pixel-level and instance-level annotations Computer Vision Civil Engineering Image Segmentation Quality Control Construction Management Concrete Technology Surface Crack Deep Learning Mirjalili, Alireza Dataset doi:10.17632/r8zrzk5w8j.2 oai:data.mendeley.com/r8zrzk5w8j.2 Creative Commons Attribution 4.0 International http://creativecommons.org/licenses/by/4.0 https://data.mendeley.com/datasets/r8zrzk5w8j 2026-09-15T09:04:40Z