文档介绍：: .
储能科学与技对采集到的锂
电池极片涂布图像进行图像预处理操作；接着，将图像中的缺陷目标区域分割出来后，提取其形态、灰度、纹
理特征；然后，搭建误差反向传播网络（Back Propagation Neural Network，BPNN），并将串行融合后的融合
特征向量作为网络的输入；最后，在训练神经网络分类模型的过程中，使用鲸鱼优化算法（Whale Optimization
Algorithm，WOA）用于辅助调参，以进一步提高模型的识别准确率。本文算法可精确实现对划痕、漏金属、孔
洞、裂纹、异污、脱碳等 8 种常见的锂电池极片涂布缺陷的检测与识别，实验结果证明，当检测的锂电池极片
宽度为 200 mm，检测精度为 mm，检测速度为 60 m·min-1 时，本文算法的平均漏检率为 %，平均误
检率为 0%，平均分类识别准确率为 %。本文算法能够有效应用于高速高精度的锂电池极片涂布缺陷检测
场合，在锂电池智能制造领域具有一定的实用价值。
关键词：机器视觉；缺陷检测；图像分类；锂电池极片
doi： .2095-
中图分类号：；TM912 文献标志码：A 文章编号：2095-4239（XXXX）XX-1-09
Defects detection and recognition of lithium battery electrode
plate coating based on WOA-BPNN
Zhong Jianping, Fei Tao
(1School of Automation Science and Engineering，South China University of Technology，Guangzhou 510641，
Guangdong，China；2Engineering Research Center for Precision Electronic Manufacturing Equipment, Ministry of
Education & Guangdong Provincial Engineering Laboratory for Advanced Chip Intelligent Packaging Equipment，Guangzhou
510641，Guangdong，China)
Abstract: The positive and negative electrode of lithium battery are important components of
lithium battery. The quality of electrode coating is largely related to the performance and
service life of the battery, and the defective electrode is often the root of the potential safety
hazard of the battery. In order to further improve the automation level of defect detection and
recognition for lithium batter