食品与机械2025，Vol.41Issue(4)：236-241,6.DOI:10.13652/j.spjx.1003.5788.2025.60008
基于改进YOLOv10的盒装食品生产线包装缺陷检测方法
Packaging defect detection method for boxed food production line based on improved YOLOv10
摘要
Abstract
[Objective]To address the problems of low detection accuracy and poor real-time performance in traditional packaging defect detection methods in food production.[Methods]Based on the analysis of automated food production lines,an improved YOLOv10 model was applied for packaging defect detection on boxed food production lines.The convolutional block attention module(CBAM)was introduced after the C2f modules in the backbone and neck layers to enhance the model's feature localization ability in complex backgrounds.Full-dimensional dynamic convolution was introduced in the backbone and neck layers to reduce computational redundancy and improve detection accuracy.The P2 layer was added and the P5 layer removed in the head to enhance small object detection performance.The loss function in the head was optimized to improve the model's convergence performance.[Results]The proposed method effectively improved the detection accuracy of packaging defects in boxed food,met real-time performance requirements,achieved a detection accuracy greater than 98.00%,and an average detection time of less than 0.02 s.[Conclusion]The integration of deep learning and machine vision can realize rapid and accurate detection of packaging defects in boxed food.关键词
食品生产/包装缺陷/YOLOv10模型/注意力机制CBAM/全维动态卷积Key words
food production/packaging defects/YOLOv10 model/convolutional block attention module/omni-dimensional dynamic convolution引用本文复制引用
王静蕾,赵英杰,李峰,王春雨..基于改进YOLOv10的盒装食品生产线包装缺陷检测方法[J].食品与机械,2025,41(4):236-241,6.基金项目
河南省科技攻关项目(编号:23010210317) （编号:23010210317）
河南省高等学校重点科研项目(编号:23A1201409) （编号:23A1201409）
