地铁隧道结构病害智能巡检机器人病害识别数据集

针对现有机器人检测效率不高、主动辨识病害能力缺失等问题，重点突破多感官机器人环境共融自主巡检机器人，研制地铁隧道结构病害智能巡检机器人。试验验证机器人的病害（裂缝、渗漏水、剥落、错台、变形、限界）自动识别类型、自动识别精度、巡检速度进行测试，记录了采集的病害图像、数据，并利用分析软件对病害识别类型、识别精度进行结果输出，对试验结果进行分析评价。

Aiming at the issues such as low detection efficiency and lack of active defect identification capability of existing robots, this study focuses on breaking through the multi-sensory environment-cooperative autonomous inspection robot technology, and develops an intelligent inspection robot for subway tunnel structural defects. Tests were conducted to evaluate the automatic recognition categories, recognition accuracy and inspection speed of the robot when identifying various defects including cracks, water seepage and leakage, spalling, misalignment, deformation and clearance limits. The collected defect images and related data were recorded, and the analysis software was used to output the results of defect recognition categories and recognition accuracy, followed by analysis and evaluation of the test results.

搜集汇总

背景与挑战

背景概述

该数据集专注于地铁隧道结构病害的智能识别，旨在通过智能巡检机器人自动检测和识别裂缝、渗漏水、剥落、错台、变形和限界等病害类型，并评估识别精度和巡检速度。数据集包含477.32MB的病害图像和数据，共176个文件，由刘新根创建，作为国家重点研发计划项目的一部分，发布于2022年，适用于隧道结构健康监测和机器人巡检技术研究。