6 December 2023 Defect analysis of cementing lenses and parameter optimization based on a convolutional neural network algorithm
Yu-Zhen Mao, Chin-Ting Ho, Chao-Hsuan Kuo, Chun-Wei Liu
Author Affiliations +
Abstract

In this research, a high-power ultraviolet-light-emitting diode was employed as a substitute for a conventional mercury lamp and found to result in a considerable reduction of the lens cementing manufacturing time from 24 h to just a few minutes. The widely recognized VGG19 architecture was employed to effectively classify images depicting cementing defects and discovered to achieve an impressive success rate of 87.26%. The training outcomes were successfully applied in subsequent experiments, which led to a marked decrease in the occurrence of cementing defects from 17.71% to a mere 1.82%. Consequently, the overall efficiency of the entire process was substantially enhanced, ultimately leading to a significant improvement in cementing lens production.

© 2023 Society of Photo-Optical Instrumentation Engineers (SPIE)
Yu-Zhen Mao, Chin-Ting Ho, Chao-Hsuan Kuo, and Chun-Wei Liu "Defect analysis of cementing lenses and parameter optimization based on a convolutional neural network algorithm," Optical Engineering 62(12), 125101 (6 December 2023). https://doi.org/10.1117/1.OE.62.12.125101
Received: 20 July 2023; Accepted: 17 November 2023; Published: 6 December 2023
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Adhesives

Cements

Lenses

Ultraviolet radiation

Optical engineering

Air contamination

Bubbles

Back to Top