Kelompok200OK/dataset_defect_detection_200_OK
收藏Hugging Face2025-07-10 更新2025-11-01 收录
下载链接：
https://hf-mirror.com/datasets/Kelompok200OK/dataset_defect_detection_200_OK
下载链接
链接失效反馈官方服务：
资源简介：
这是一个真实的工业铸造产品数据集，用于制作预测模型以检测潜水电泵叶轮的缺陷。数据集中的图片尺寸为512x512像素，且为灰度图像。该数据集未进行进一步的数据增强处理，包含1300张图片，其中519张标记为正面无缺陷（ok_front），781张标记为正面有缺陷（def_front）。数据集被划分为训练集和验证集，分别包含1040张和260张图片。
This is a real-life industrial casting product dataset intended for building a predictive model to detect defects in submersible pump impellers. The images in the dataset are 512x512 pixels in grayscale. The dataset has not undergone further data augmentation and consists of 1300 images, with 519 labeled as front without defects (ok_front) and 781 labeled as front with defects (def_front). The dataset is divided into a training set and a validation set, containing 1040 and 260 images respectively.
提供机构：
Kelompok200OK
