摘要
在电池制造业中,缺陷检测技术已成为确保电池性能、安全性和成本效益的关键环节。文中首先针对二维图像处理方法,总结了不同的图像分割及特征提取方法。接着介绍了基于机器学习的动力电池缺陷检测领域的研究进展,归纳了在已知缺陷类型、未知缺陷类型和少量缺陷类型等不同类型在电池缺陷检测中的研究现状和取得的主要成果。最后,基于当前动力电池缺陷检测面临的技术瓶颈,对未来研究方向进行了展望。
In the battery manufacturing industry,defect detection technology has emerged as a critical link in ensuring battery performance,safety,and cost-effectiveness.In this paper,different image segmentation and feature extraction methods for two-dimensional image processing are initially summarized.Subsequently,the research advancements in the field of power battery defect detection via machine learning are presented,and the research status and principal achievements in various types of battery defect detection,including known defect types,unknown defect types,and a small number of defect types,are generalized.Finally,based on the technical bottlenecks in current power battery defect detection,the future research directions were prospected.
作者
墙明星
王权
QIANG Mingxing;WANG Quan(School of Mechanical Engineering,Jiangsu University)
出处
《仪表技术与传感器》
北大核心
2025年第8期73-79,共7页
Instrument Technique and Sensor
关键词
动力电池
缺陷检测
二维图像处理
机器学习
深度学习
power battery
defect detection
two-dimensional image processing
machine learning
deep learning
作者简介
墙明星(1996-),硕士研究生,主要从事机器视觉工业缺陷检测方法研究。E-mail:mingxing_mail@163.com;通信作者:王权(1973-),教授,博士,主要从事半导体传感技术和机器视觉等研究。E-mail:wangq@ujs.edu.cn。
