Surface pit defects detecting method of pouch Li-ion battery
The traditional method is difficult to accurately detect the surface pit defects of pouch Li-ion battery due to their low contrast,small defect area and reflection.A method for detecting the pit defect on the surface of the pouch Li-ion battery based on image enhancement and improved DeepLabV3 network is proposed.The image features of pit defects on the surface are analyzed,the image enhancement algorithm is used to preprocess the image to enhance the contrast of pit defects.The DeepLabV3 network is improved,ResNet101 is used as the feature extraction network.The position attention module is introduced to make the model pay more attention to the related features of pit defects and improve the detection accuracy of the network.The median intersection-over-union of the improved network on the self-made data set reaches 85.98%,the accuracy of defect detection reaches 98.33%.
万方数据
维普
