Pavement cracks from UAV imagery-1388

A pavement crack image dataset from UAV imagery integrated with <b>GSD</b> was established and has been made publicly available for the research community, serving as a valuable supplement to existing crack databases.A total of 1388 pavement crack images here were collected and labeled, with 304 samples identified as being of the longitudinal crack (<b>LC</b>) type, 303 samples identified as being of the transverse crack (<b>TC</b>) type, 313 samples identified as being of the obliquely oriented crack (<b>OC</b>) type, 368 samples identified as being of the alligator crack (<b>AC</b>) type, and 100 samples being identified as of the no-crack type. To ensure the deep learning-based model’s effectiveness, the datasets were divided into training, validation, and test sets in the ratio of 80%, 10%, and 10%, respectively.Please cited this reference as following if using the data.XinbaoChenX; Chang Liu; Long Chen; Xiaodong Zhu; Yaohui Zhang; Chenxi Wang ; A PavementCrack Detection and Evaluation Framework for a UAV Inspection System Based on Deep Learning,<i>Applied Sciences</i>, 2024, 14(3): 1157.<br>

本研究构建了一套融合地面采样距离（GSD）的无人机航拍路面裂缝图像数据集，并面向科研社区公开共享，作为现有裂缝数据库的有益补充。本次数据集共采集并标注了1388张路面裂缝图像，其中纵向裂缝（LC）样本304张、横向裂缝（TC）样本303张、斜向裂缝（OC）样本313张、龟裂（AC）样本368张，另有无裂缝样本100张。为保障基于深度学习的模型训练效果，本数据集按照80%、10%、10%的比例划分为训练集、验证集与测试集。若使用本数据集，请按以下格式引用该文献：陈新宝、刘畅、陈龙、朱晓东、张耀辉、王晨曦. 基于深度学习的无人机巡检系统路面裂缝检测与评估框架[J]. 应用科学（Applied Sciences）, 2024, 14(3): 1157.

搜集汇总

背景与挑战

背景概述

该数据集包含1388张无人机拍摄的已标注路面裂缝图像，涵盖四种裂缝类型和无裂缝样本，并按8:1:1比例划分为训练/验证/测试集，为基于深度学习的路面裂缝检测研究提供了补充数据。