基于机器视觉的食品外包装缺陷检测算法研究进展

食品包装在生产过程中由于各种因素会导致缺陷产生，包装缺陷种类多，背景复杂。通过视觉成像和计算机信息处理完成包装的识别、检测和测量等任务的机器视觉检测，相比传统的人工检测，具有执行速度快、精度高等特点，可显著提高生产自动化程度。文章根据食品外包装常见缺陷，从缺陷检测算法的角度介绍传统机器视觉检测算法和深度学习相关算法在食品外包装缺陷检测中的研究应用，并对检测算法在食品外包装缺陷检测中的应用前景，以及存在的问题进行分析与展望。

Research progress of food packaging defect detection based on machine vision

Food packaging can develop defects during the production process due to various factors. The types of packaging defects are numerous with complex background. Machine vision detection, which uses visual imaging and computer information processing to complete tasks such as identification, detection, and measurement of packaging, has faster execution speed and higher accuracy compared to traditional manual inspection. This can significantly improve the degree of production automation. This article analyzes the common defects in food packaging and their causes, introduces traditional machine vision detection algorithms, and explores the research application of deep learning algorithms in food packaging defect detection. It also analyzes the prospects and challenges of applying detection algorithms in food packaging defect detection.