Fabric Defect Detection Based on Improved UNet Model and Transfer Learning
To solve the problems of complex algorithm design,low detection speed and large model size in traditional fabric de-fect detection methods,we propose a fabric defect detection method based on improved UNet model and transfer learning,which uti-lizes the improved UNet model as the framework to construct the fabric defect detection model.The model size and computational complexity are significantly reduced by reducing the number of channels in the dual convolution module of the UNet model.Mean-while,the convergence speed and detection ability of the model are enhanced by adopting transfer learning strategy and variable learning rates.The experimental results show that the proposed method achieves good performance in fabric defect detection tasks.Moreover,the proposed method has the advantages such as high detection speed,small model size,and no need for large-scale train-ing samples.The proposed method effectively addresses the problems of traditional methods and facilitates deployment and applica-tion in real-world industrial scenarios.
NETL
NSTL
万方数据
