Detection Method of Surface Defects of Electronic Components Using the Deep Network of You Only Look Once Version 5

Konferenz: MEMAT 2022 - 2nd International Conference on Mechanical Engineering, Intelligent Manufacturing and Automation Technology
07.01.2022 - 09.01.2022 in Guilin, China

Tagungsband: MEMAT 2022

Seiten: 5Sprache: EnglischTyp: PDF

Autoren:
Lin, Yihao; Fan, Yong; Gong, Yubing (School of Mechanical & Electrical Engineering, Guilin University of Electronic Technology, Guilin, Guangxi, China)
Wu, Shuqin (School of Architecture and Transportation Engineering, Guilin University of Electronic Technology, Guilin, Guangxi, China)
Xie, Wu (School of Computer Science and Information Security, Guilin University of Electronic Technology, Guilin, Guangxi, China)

Inhalt:
The production of electronic components is becoming more and more modern and efficient. How to detect surface defects of electronic components more accurately and quickly has become a more and more important research topic. To solve this problem, this paper designs a detection method based on visual detection and image processing is proposed. The proposed method builds a target detection algorithm model based on the YOLOv5 (You Only Look Once Version 5) convolutional neural network, and combines the image target detection and recognition with the convolutional neural network to make its detection and recognition effect more accurate and efficient. This paper tested the 4 surface defects with the patch resistance SS14 to verify the reliability of the system. Through the result analysis, the loss function of the detection system is less than 0.3, and its map is above 0.96. It meets the high accuracy and high accuracy requirements of the surface defect detection of electronic components.