Author:
Cai Zhishan,Chen Jiangwei,Yan Shaobin,Li Shizhan,Liao Tingdi
Abstract
Abstract
Technical schemes such as uniform TEC component feeding, posture adjustment, multi-station positioning, and communication between upper and lower machines were designed. The opto-electromechanical integration mechanism was integrated by software to realize the detection of TEC component surface defects, and qualified TEC components were separated from unqualified TEC components by blow sieving. Adopt ARM MCU for real-time control; Image recognition adopts the machine vision deep learning platform based on linux to learn and recognize the TEC components, and sends the results to the lower computer for sieving. The accuracy of triggering the camera to take pictures is 99.9%, which provides a guarantee for capturing and processing images accurately. The accuracy of blow sieving is over 99.9%, which can meet the sieving requirements. The electrical control system of the whole machine can meet the motion control requirements of detection.
Subject
General Physics and Astronomy
Reference10 articles.
1. A survey of surface defect detection methods based on deep learning;Tao;Acta Automatica Sinica,2021
2. Chip resistance direction detection system based on Halcon;Kang;Packaging Engineerin,2017
Cited by
1 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献
1. Research on Intelligent Network Deception Technology for Attack Trapping;2023 International Conference on Computer Simulation and Modeling, Information Security (CSMIS);2023-11-15