lipi17/building-cracks
收藏资源简介:
该数据集通过Roboflow.com于2023年2月10日下午3:51 GMT导出。Roboflow是一个端到端的计算机视觉平台,帮助用户在计算机视觉项目中与团队协作、收集和组织图像、理解和搜索非结构化图像数据、注释和创建数据集、导出、训练和部署计算机视觉模型,并使用主动学习来改进数据集。数据集包括2134张图像,土壤以COCO格式进行注释。每张图像都应用了以下预处理步骤:自动定向像素数据(去除EXIF方向信息)和调整大小为640x640(拉伸)。未应用任何图像增强技术。
This dataset was exported via Roboflow.com at 3:51 PM GMT on February 10, 2023. Roboflow is an end-to-end computer vision platform that enables users to collaborate with teams, collect and organize images, comprehend and search unstructured image data, annotate and create datasets, export, train and deploy computer vision models, and utilize active learning to enhance datasets. This dataset includes 2134 images, with annotations for soil formatted in COCO format. The following preprocessing steps were applied to each image: automatic orientation of pixel data (removal of EXIF orientation metadata) and resizing to 640×640 via stretching. No image augmentation techniques were applied.
数据集概述
任务类别
- 目标检测
标签
crack
图像数量
- 训练集:1490张
- 验证集:433张
- 测试集:211张
使用方法
-
安装
datasets库: bash pip install datasets -
加载数据集: python from datasets import load_dataset
ds = load_dataset("lipi17/building-cracks", name="full") example = ds[train][0]
引用
@misc{ crack-detection-y5kyg_dataset, title = { Crack Detection Dataset }, type = { Open Source Dataset }, author = { António Raimundo }, howpublished = { url{ https://universe.roboflow.com/antonio-raimundo/crack-detection-y5kyg } }, url = { https://universe.roboflow.com/antonio-raimundo/crack-detection-y5kyg }, journal = { Roboflow Universe }, publisher = { Roboflow }, year = { 2023 }, month = { feb }, note = { visited on 2023-10-21 }, }
许可证
CC BY 4.0
数据集摘要
- 数据集通过 roboflow.com 导出,导出时间为2023年2月10日 15:51 GMT。
- 数据集包含2134张图像。
- 图像标注格式为COCO。
- 图像预处理包括自动方向校正和调整大小至640x640(拉伸)。
- 未应用图像增强技术。




