Coal mine underground drilling site object detection dataset

Drilling in underground coal mine is an important measure to solve gas disaster, water disaster and hidden geological disaster, which can significantly improve the level of coal mine disaster prevention and control. In order to monitor the drilling process in real time and improve drilling efficiency, object detection in coal mine underground drilling site is required. Object detection in coal mine underground drilling site is to identify and locate important targets involved in the drilling site. Compared with the traditional coal mine drilling site object detection method, the deep learning-based coal mine drilling site object detection method can improve the accuracy, timeliness and stability of the object detection, but it needs to rely on high-quality dataset. At present, research on object detection in coal mine underground drilling site mainly relies on small-scale private dataset, which is difficult to provide sufficient and reliable data for deep neural network model training. This study uses intrinsically safe law enforcement recorders for coal mine to photograph coal mine underground drilling site, and through steps such as data cleaning, data annotation, and expert spot inspection, a standardized object detection dataset for coal mine underground drilling site is constructed. This dataset contains 70,948 images from different drilling sites and environmental background conditions, covering five categories of objects: gripper, chuck, coal miner, mine safety helmet, and drill pipe, and provides annotated files in PASCAL VOC format. This dataset can provide strong data support for object detection research in coal mine underground drilling site, and plays an important role in promoting intelligent coal mine underground monitoring and early warning.地下煤矿钻探是解决瓦斯灾害、水害及隐蔽地质灾害的重要手段，可显著提升煤矿灾害防控水平。为实时监测钻探过程、提升钻探效率，需开展煤矿井下钻探现场目标检测任务。煤矿井下钻探现场目标检测，即识别并定位钻探现场涉及的各类重要目标。相较于传统煤矿钻探现场目标检测方法，基于深度学习的煤矿钻探现场目标检测方法可提升检测的精准度、时效性与稳定性，但需依赖高质量数据集。当前，煤矿井下钻探现场目标检测相关研究多依赖小型私有数据集，难以满足深度神经网络模型训练对充足可靠数据的需求。本研究采用煤矿用本安型执法记录仪拍摄煤矿井下钻探现场，经数据清洗、数据标注、专家现场核验等流程，构建了标准化的煤矿井下钻探现场目标检测数据集。该数据集包含来自不同钻探现场及环境背景条件下的70948张图像，涵盖夹持器（gripper）、卡盘（chuck）、煤矿工人、矿工安全帽及钻杆（drill pipe）五类目标，并提供PASCAL VOC格式的标注文件。该数据集可为煤矿井下钻探现场目标检测研究提供强有力的数据支撑，对推动煤矿井下智能监测预警具有重要意义。

搜集汇总

背景与挑战

背景概述

该数据集是一个专注于煤矿井下钻场目标检测的大规模图像数据集，包含70,948张图像，覆盖夹持器、卡盘、矿工、矿用安全帽和钻杆五类对象，并提供PASCAL VOC格式的标注文件，旨在支持深度学习模型训练，提升煤矿井下监测的智能化水平。数据集由西安科技大学团队使用本质安全执法记录仪构建，经过严格的数据处理流程，数据量为8.80 GB，适用于煤矿工程和计算机科学领域的研究。