实验室穿戴识别合规与人流统计数据集

本数据集计算机端以图片、视频为原始数据载体，经标注标框处理后，生成JSON格式结构化数据进行存储。数据组织采用“单数据对应单文件”模式，每张/段音视频数据对应唯一JSON文件，文件内整合原始数据存储路径、标框坐标信息及核心标注字段结果，实现数据关联清晰、检索高效。
核心标注字段及属性均为布尔型（取值为“是/否”），具体如下：wear labcoat（穿实验服），标识目标人员是否穿戴实验服；no labcoat（未穿实验服），标识目标人员是否未穿戴实验服；wear goggles（戴护目镜），标识目标人员是否佩戴护目镜；no goggles（未戴护目镜），标识目标人员是否未佩戴护目镜；wear gloves（戴手套），标识目标人员是否佩戴防护手套；no gloves（未戴手套），标识目标人员是否未佩戴防护手套。
主要关键字段为标框坐标（关联视觉区域）与各穿戴合规字段（核心业务属性），通过关键字段可快速检索对应视觉区域的穿戴合规状态，为实验室穿戴识别的数据分析与模型训练提供支撑。

This dataset uses images and videos as the original data carriers for computer-based applications. After bounding box annotation processing, it generates JSON-formatted structured data for storage. The data organization adopts the "one data entry per single file" mode: each image or video segment corresponds to a unique JSON file, which integrates the original data storage path, bounding box coordinate information, and core annotation field results, enabling clear data association and efficient retrieval.

All core annotation fields and their attributes are of boolean type (with values of "yes" or "no"), as detailed below:
1. wear labcoat: indicates whether the target personnel is wearing a lab coat;
2. no labcoat: indicates whether the target personnel is not wearing a lab coat;
3. wear goggles: indicates whether the target personnel is wearing protective goggles;
4. no goggles: indicates whether the target personnel is not wearing protective goggles;
5. wear gloves: indicates whether the target personnel is wearing protective gloves;
6. no gloves: indicates whether the target personnel is not wearing protective gloves.

The main key fields are the bounding box coordinates (associated with visual regions) and each wearing compliance field (core business attributes). These key fields allow for rapid retrieval of the wearing compliance status of the corresponding visual regions, providing support for data analysis and model training in laboratory personal protective equipment (PPE) wearing recognition.

搜集汇总

背景与挑战

背景概述

该数据集旨在通过视觉识别技术，对实验室环境内的人员穿戴合规性进行检测与统计，并可能同步实现人流量监测。数据集预期包含实验室场景下的图像或视频数据，标注信息应涵盖安全防护装备穿戴状态、人员位置及计数等，适用于实验室安全管理与行为分析模型的开发。