Classification of Solder Quality in Through-Hole Devices by Convolutional Neural Networks Evrisimsel Sinir Aglari ile Through-Hole Cihazlarda Lehim Kalitesinin Siniflandirilmasi


Sarigul N., YILDIRIM T.

2022 Innovations in Intelligent Systems and Applications Conference, ASYU 2022, Antalya, Turkey, 7 - 09 September 2022 identifier

  • Publication Type: Conference Paper / Full Text
  • Doi Number: 10.1109/asyu56188.2022.9925424
  • City: Antalya
  • Country: Turkey
  • Keywords: automated optical inspection, electronics production, Inception-ResNet, Inception-v4, ResNet, soldering
  • Yıldız Technical University Affiliated: Yes

Abstract

Solder inspection has an important place in electronic board production. Deep learning methods can be used to control the quality of solders. In this paper, a comparative study on the performance of deep learning methods for the classification of solder quality of through-hole devices is presented. The dataset used contains 7320 pieces of data. Soldering quality will be classified using three different networks based on ResNet, Inception-v4 and Inception-ResNet, which were created in this study. The test results obtained will be compared.