铝型材缺陷检测数据集APDDD

本数据集图片来源于2018广东工业智造大数据创新大赛-智能算法赛《铝型材表面瑕疵识别》的初赛开源数据集：https://tianchi.aliyun.com/competition/entrance/231682/information。

该比赛初赛为分类比赛，对应数据集仅有分类标签，我对原分类数据集进行缺陷标注，将其重构为缺陷检测数据集，标注后的检测数据集命名为APDDD(Aluminum Profile Surface Detection Database)。

The images of this dataset are sourced from the preliminary round open-source dataset of the "Aluminum Profile Surface Defect Recognition" track in the 2018 Guangdong Industrial Intelligent Manufacturing Big Data Innovation Competition, accessible at: https://tianchi.aliyun.com/competition/entrance/231682/information. The preliminary round of this competition was a classification task, and the original corresponding dataset only contained classification labels. I conducted defect annotation on the original classification dataset and reconstructed it into a defect detection dataset. The annotated detection dataset is named APDDD (Aluminum Profile Surface Detection Database).

搜集汇总

背景与挑战

背景概述

该数据集是铝型材表面缺陷检测数据集APDDD，来源于2018年广东工业智造大数据创新大赛的初赛开源数据，经过重构和缺陷标注从分类任务转换为检测任务。数据集包含2560×1920尺寸的图片，涵盖十类常见缺陷，如凹陷、擦花和涂层开裂等，适用于工业视觉检测应用。