安全头盔佩戴检测图像数据集

当前目录为“helmet_dataset”，内含用于安全头盔佩戴检测的图片数据及标签。目录下主要包含两个子文件夹“images”、“labels”以及配置文件“data.yaml”。
“images”文件夹下设“train”、“val”、“test”三个子集，分别存放379张、81张和80张.jpg格式的图片，内容均为是否佩戴安全头盔的场景。
“labels”文件夹下设同名的“train”、“val”、“test”子目录，各包含与图片数量对应的.txt标签文件。每个.txt文件与同名的图片文件一一对应，文件内包含一行数据，由5个用空格分隔的数字组成：首位为类别标签（0表示未佩戴，1表示佩戴），第2-3位为目标中心点的归一化坐标，第4-5位为目标的归一化宽高。
该数据结构符合YOLO模型训练格式，适用于目标检测任务。

The current working directory is "helmet_dataset", which contains image data and labels for safety helmet wearing detection. The directory mainly includes two subfolders "images", "labels" and a configuration file "data.yaml".
The "images" folder contains three subsets: "train", "val", and "test", which store 379, 81, and 80 .jpg format images respectively, all depicting scenarios related to whether safety helmets are worn or not.
The "labels" folder has subdirectories with the same names: "train", "val", and "test", each containing .txt label files matching the number of corresponding images. Each .txt file has a one-to-one correspondence with the image file of the same name, and contains a single line of data composed of 5 space-separated numbers: the first value is the category label (0 represents not wearing a helmet, 1 represents wearing a helmet), the 2nd and 3rd values are the normalized coordinates of the target's center point, and the 4th and 5th values are the normalized width and height of the target.
This data structure complies with the training format of YOLO models and is applicable to object detection tasks.

搜集汇总

背景与挑战

背景概述

该数据集名称为'安全头盔佩戴检测图像数据集'，但提供的详情内容显示为上海市数据产品知识产权登记公示列表，主要包含非遗文化类三维模型产品的登记信息，如绳编、钩针、根雕等数字化模型。由于内容与数据集名称不符，无法直接总结该数据集的具体信息，建议参考其他来源获取准确描述。