一种基于卷积神经网络的轻量级焊缝缺陷识别算法

打开文本图片集

关键词：焊缝缺陷;卷积神经网络;MobileNetV3;fire模块;ECA模块;轻量级;智能识别 中图分类号：TP391.14 文献标志码：A doi：10.12415/j.issn.1671-7872.23178

A Lightweight Weld Defect Identification Algorithm Based on Convolutional Neural Network

XU Feihu， LI Dan， ZHAO Wenjie

（School of Electrical & Information Engineering， Anhui University of Technology，Maanshan ， China）

Abstract： To address the issues of low recognition eficiency and accuracy in traditional weld defect detection methodswhen processing large-scale industrial data，a lightweight weld defect recognition algorithm based on convolutional neural networks was proposed.The fire module was introduced into the original MobileNetV3 to reduce parameter size，whilethe ECA （eficient channel atention） module was incorporated to enhance feature chanel learming capability，thereby optimizing computational resource allocation and improving feature extraction performance.To validate the efectiveness of the proposed algorithm，comparative experiments were conducted with common classification models on a weld defect test dataset. The results demonstrate that，compared to other classification models， an average recognition accuracy of 98.50% is achieved by the proposed algorithm for common industrial defects such as dents，pores，and burrs，with the original algorithm being significantly outperformed， thanks to the combined effects of the fire module's lightweight design and the ECA module's feature enhancement. Moreover，both parameter size and floating-point operationsaresignificantly reduced by the improved MobileNetV3 algorithm while high recognitionaccuracy is maintained，making it particularlysuitable for deployment on industrial inspection devices with limited computational resources.A practical solution is thus provided for real-time quality inspection in the field of intelligent manufacturing.

Keywords： weld seam defect; convolutional neural networks; MobileNetV3;fire module; ECA module; lightweight; intelligent identify

焊接工艺作为一种通过加热、高温或高压实现金属和热塑性材料接合的关键制造技术，在现代工业的铁路建设、桥梁工程、汽车制造、航空航天等核心领域具有不可替代的应用价值。（剩余12679字）