DanielCerda/pid-object-detection
收藏资源简介:
该数据集用于目标检测任务,包含5种标签:球阀、蝶阀、离心泵、止回阀和闸阀。数据集总共有152张图像,分为训练集(128张)、验证集(12张)和测试集(12张)。图像以COCO格式标注,未应用任何图像增强技术。数据集通过Roboflow平台导出,发布于2023年2月10日,遵循CC BY 4.0许可。
This dataset is intended for object detection tasks, and includes 5 types of labels: ball valve, butterfly valve, centrifugal pump, check valve, and gate valve. In total, it contains 152 images, which are split into the training set (128 images), validation set (12 images), and test set (12 images). All images are annotated in COCO format, and no image augmentation techniques have been applied. The dataset was exported via the Roboflow platform, released on February 10, 2023, and is licensed under CC BY 4.0.
数据集概述
任务类别
- 目标检测
标签
- 球阀(ball-valve)
- 蝶阀(butterfly-valve)
- 离心泵(centrifugal-pump)
- 止回阀(check-valve)
- 闸阀(gate-valve)
图像数量
- 训练集:128张
- 验证集:12张
- 测试集:12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("DanielCerda/pid-object-detection", name="full") example = ds[train][0]
引用
@misc{ pid_dataset_dataset, title = { pid_dataset Dataset }, type = { Open Source Dataset }, author = { PID Smart Reader }, howpublished = { url{ https://universe.roboflow.com/pid-smart-reader/pid_dataset } }, url = { https://universe.roboflow.com/pid-smart-reader/pid_dataset }, journal = { Roboflow Universe }, publisher = { Roboflow }, year = { 2023 }, month = { feb }, note = { visited on 2023-10-28 }, }
许可证
CC BY 4.0
数据集摘要
- 数据集通过 roboflow.com 于2023年2月10日3:14 AM GMT 导出。
- 包含152张图像。
- 管道元件以COCO格式标注。
- 未应用图像增强技术。




