Research on Surface Damage Classification Algorithm of Positive Temperature Coefficient Thermistor Based on Transfer Learning
KE Yu
Abstract：Positive temperature coefficient resistors(PTCR)are susceptible to surface damage during transfer and transporta-tion in the manufacturing process,so they are usually screened for damage before they leave the factory.However,existing methods often use manual sorting,which is inefficient and easy to miss,so it is especially important to use machine vision to perform intelli-gent sorting.In this paper,a migration learning-based algorithm for classifying PTCR surface damage is proposed.This algorithm ex-tracts feature vectors in three convolutional neural network models Inception-V3,ResNet50,and Xception after data enhancement of the acquired dataset,and then combines the extracted feature vectors for training and uses SoftMax for classification,and finally obtains OurNet(self-named)model with the highest accuracy.The model can automatically identify three kinds of surface damage,namely,edge breakage,complete rupture and surface scratching,so as to guarantee product quality and improve production effi-ciency.
Keywords：PTCRdamage classificationtransfer learning
Publication Date：2024-06-20
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：5( 1703-1707 )