http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
직렬형 컨볼루션 신경망에 의한 SMD 부품의 조립 불량 분류 방법
류종현(Jong-Hyun Ryu),김영규(Young-Gyu Kim),박태형(Tae-Hyoung Park) 대한전기학회 2019 전기학회논문지 Vol.68 No.10
In this paper, we propose the classification method of assembly defects in the surface mount technology process. We used a cascade convolution neural network which two convolutional neural networks were merged into one network, for assembly defect classification. The first network classifies whether the surface mount device (SMD) is defect or not. The second network classify the result of the first network more detail. We classified the SMD defects as six types using a cascade convolution neural network. Experiment result shows that the proposed method can optimize memory usage and improve classification accuracy compared to previous methods.