Rapid Method Using Deep Learning with Multi-focus Microphotographs to Measure Submicrometric Structures and Its Evaluation

Confocal laser scanning microscopy (CLSM) and scanning electron microscope (SEM) systems are commonly used for measuring the dimensions of laser-processed objects. Nevertheless, both methods require some time for preprocessing and measurement, thereby entailing high costs. We propose a simple, fast,...

Full description

Saved in:
Bibliographic Details
Published in:Journal of laser micro nanoengineering 2021-10, Vol.16 (2), p.150-154
Main Authors: Narukage, Riki, Okada, George, Kawaguchi, Hiroshi
Format: Article
Language:English
Subjects:
Accuracy
Deep learning
Experiments
Grooves
Integrated circuit fabrication
Laser applications
Lasers
Machine learning
Manufacturing
Measurement
Measurement methods
Methods
Microphotographs
Microscopy
Neural networks
Normal distribution
Optical microscopes
Scanning microscopy
Semantics
Silicon wafers
Technology application
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Confocal laser scanning microscopy (CLSM) and scanning electron microscope (SEM) systems are commonly used for measuring the dimensions of laser-processed objects. Nevertheless, both methods require some time for preprocessing and measurement, thereby entailing high costs. We propose a simple, fast, and inexpensive method for measuring submicrometric structures using deep learning with multi-focus microphotographs taken using an optical microscope. The average errors in depth and height for a laser-processed groove and a laser-processed ridge are, respectively, 0.1667 [micro]m and 0.4349 [micro]m. The estimation time is 971.50 ms for the 64 * 64 [micro][m.sup.2] area. Keywords: machine learning, microphotograph, submicrometric structure measurement
ISSN:1880-0688
1880-0688
DOI:10.2961/jlmn.2021.02.3001