Share Email Print

Proceedings Paper

Deep learning for evaluating difficult-to-detect incomplete repairs of high fluence laser optics at the National Ignition Facility
Author(s): T. Nathan Mundhenk; Laura M. Kegelmeyer; Scott K. Trummer
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

Two machine-learning methods were evaluated to help automate the quality control process for mitigating damage sites on laser optics. The mitigation is a cone-like structure etched into locations on large optics that have been chipped by the high fluence (energy per unit area) laser light. Sometimes the repair leaves a difficult to detect remnant of the damage that needs to be addressed before the optic can be placed back on the beam line. We would like to be able to automatically detect these remnants. We try Deep Learning (convolutional neural networks using features autogenerated from large stores of labeled data, like ImageNet) and find it outperforms ensembles of decision trees (using custom-built features) in finding these subtle, rare, incomplete repairs of damage. We also implemented an unsupervised method for helping operators visualize where the network has spotted problems. This is done by projecting the credit for the result backwards onto the input image. This shows regions in an image most responsible for the networks decision. This can also be used to help understand the black box decisions the network is making and potentially improve the training process.

Paper Details

Date Published: 14 May 2017
PDF: 8 pages
Proc. SPIE 10338, Thirteenth International Conference on Quality Control by Artificial Vision 2017, 103380H (14 May 2017); doi: 10.1117/12.2264000
Show Author Affiliations
T. Nathan Mundhenk, Lawrence Livermore National Lab. (United States)
Laura M. Kegelmeyer, Lawrence Livermore National Lab. (United States)
Scott K. Trummer, Lawrence Livermore National Lab. (United States)

Published in SPIE Proceedings Vol. 10338:
Thirteenth International Conference on Quality Control by Artificial Vision 2017
Hajime Nagahara; Kazunori Umeda; Atsushi Yamashita, Editor(s)

© SPIE. Terms of Use
Back to Top
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?