基于X光透射的包装食品危险异物灰度图像数据集

数据集为基于 X 光透射的包装食品危险异物灰度图像数据集，覆盖了包括金属（如铁丝、钉子）、非金属（如玻璃、塑料）、天然异物（如贝壳、石子、骨渣等）在内的多种类型异物，经工业级 X-Ray 异物检测机单模态采集，同步记录时间戳等信息，还做了数据清洗与标准化处理，并用工具标注异物位置。数据集以 X 光灰度图像为主要内容，含对应标注及食品类型等元数据，分组存储。用途上，适用于速冻食品等包装食品流水线自动化检测，可用于目标检测等模型，支持相关算法设计、模型训练及系统部署。

This is a grayscale image dataset for hazardous foreign matter detection in packaged foods based on X-ray transmission. It covers multiple types of foreign matter including metals (e.g., iron wires, nails), non-metallic materials (e.g., glass, plastics), and natural foreign substances such as shells, stones, bone scraps, etc. The dataset was collected via a single-modal industrial-grade X-ray foreign matter detector, with synchronized logging of metadata including timestamps. Preprocessing steps including data cleaning and standardization have been performed, and the positions of foreign matters have been annotated using specialized tools. The dataset primarily comprises X-ray grayscale images, paired with corresponding annotations and supplementary metadata such as food category, and is stored in grouped storage units. For application scenarios, this dataset is suitable for automated inspection on packaged food production lines such as quick-frozen foods, can be applied to models like object detection, and supports relevant algorithm design, model training and system deployment.