Fabric Defects Object Detection Dataset

The dataset comprises 720 images each depicting four distinct defect types in colored fabrics: Oil, Hole, Cutting, and Crack, totaling approximately 180 images per type. This dataset holds significant academic value, particularly within the realm of Computer Vision, serving as a crucial resource for developing image processing algorithms and deep learning models for tasks such as classification, object detection, or segmentation. Its application stands to significantly enhance advancements in textile engineering and manufacturing processes.<br>The FD_Dataset comprises two distinct files:<br><br>FD, which encompasses comprehensive datasets alongside Label.mat facilitated through MATLAB for any object detection model.YOLO, housing datasets for training, validation, and testing specifically tailored for YOLO object detection models.<br>本数据集共包含720张彩色织物瑕疵图像，涵盖四类典型瑕疵类型：油渍（Oil）、破洞（Hole）、裁剪瑕疵（Cutting）与裂纹（Crack），每类对应图像约180张。本数据集具备显著学术价值，尤其在计算机视觉（Computer Vision）领域，是开发图像处理算法与深度学习模型的核心资源，可支撑分类、目标检测、语义分割等相关任务，其应用可有效推动纺织工程与制造流程的技术进步。 FD_Dataset 包含两个独立的文件组： 其一为FD组，包含完整数据集与可通过MATLAB加载的Label.mat文件，适配各类目标检测模型；其二为YOLO组，内置专为YOLO目标检测模型定制的训练、验证与测试数据集。