基于机器视觉的食品瓶罐包装缺陷检测研究进展

根据近年文献资料，介绍了机器视觉技术在食品瓶罐包装缺陷检测上的应用和发展，概述瓶罐包装缺陷检测系统的硬件结构和检测流程；梳理并分析基于图像处理技术和基于深度学习方法在食品瓶罐包装缺陷检测领域取得的研究成果，探讨分类网络模型和目标检测网络模型两种技术在该领域的优势和不足，并对未来的发展进行了展望，为智能瓶罐包装检测的创新和发展提供参考。

Research Progress on Inspecting Food Bottle and Can Packaging Shortcomings Based on Machine Vision

Based on the literature in recent years, the application and development of machine vision in the detection of food bottle and can packaging defects were introduced, and the hardware structure and detection process of the detection system were explained. The research results based on image processing technology and deep learning method in the field of packaging defect detection of food bottles and cans were reviewed and analyzed. The research also summarized the advantages and disadvantages of classification network model and target detection network model both technologies and prospected their future development, and provide a reference for the innovative development of intelligent packaging inspection.