地铁工程安全隐患智能诊断大规模数据集

安全隐患图像的采集主要来源于人工拍照收集，辅助摄像头视频智能识别。对采集到的图片及视频进行加工处理，对标安全隐患数据库得到结构化、标准化的数据。同时对整改各环节实现数字化记录留痕，形成闭环管理数据，并根据评估模型等对隐患等级进行动态标注。

The acquisition of safety hazard images primarily relies on manual photography, with intelligent recognition of surveillance camera videos serving as a supplementary approach. The collected images and videos are processed, then aligned against the safety hazard database to produce structured and standardized datasets. Meanwhile, every stage of hazard rectification is digitally documented and tracked to form closed-loop management datasets, and the hazard levels are dynamically labeled based on evaluation models and other relevant assessment tools.

搜集汇总

背景与挑战

背景概述

该数据集是一个针对地铁工程安全隐患智能诊断的大规模数据集，包含1400条结构化数据，每周更新。数据来源于企业，通过人工和智能识别采集，应用于轨道交通工程的安全隐患智能抓取和整改。