"Mirror-Assisted Multi-View Bottle Defect Detection Dataset"

"# Mirror-Assisted Multi-View Bottle Defect Detection DatasetThis dataset contains mirror-assisted multi-view images of cylindrical industrial bottles collected for real-time defect detection and automated visual inspection research. The dataset was acquired using a synchronized multi-camera imaging setup with planar mirrors designed to reduce geometric distortion and specular reflections during industrial bottle inspection.The dataset includes annotated images across multiple defect categories such as scratches, rolling stickers, seal damage, label defects, and date-print anomalies. Annotations are provided in YOLO object detection format for deep learning-based industrial inspection research.The dataset was collected and evaluated in collaboration with Red Ginseng Buyeo Factory, South Korea. Due to confidentiality agreements and industrial data-sharing restrictions, the dataset is not publicly available and is intended strictly for internal research, evaluation, and authorized academic purposes only.Included files:- Training and validation images- YOLO annotation files- Dataset configuration files- Demonstration video of the acquisition setupResearch Areas:- Computer Vision- Industrial Inspection- Deep Learning- Object Detection- Intelligent ManufacturingUsage, redistribution, or commercial application of the dataset without permission from Red Ginseng Buyeo Factory and the authors is prohibited."