Research on feature surface thickness prediction based on transformer deep-learning-networks

Xingzuo Li; Chengzhi Su; Feng Jiang; Guang Zhao; Yufei Mei

doi:10.1117/12.3033021

5 July 2024 Research on feature surface thickness prediction based on transformer deep-learning-networks

Xingzuo Li, Chengzhi Su, Feng Jiang, Guang Zhao, Yufei Mei

Author Affiliations +

Proceedings Volume 13184, Third International Conference on Electronic Information Engineering and Data Processing (EIEDP 2024); 1318466 (2024) https://doi.org/10.1117/12.3033021
Event: 3rd International Conference on Electronic Information Engineering and Data Processing (EIEDP 2024), 2024, Kuala Lumpur, Malaysia

Abstract

Aiming at the problem of predicting the amount of material surface removal, this paper is based on the development and application of deep learning and the latest research results. It takes images of paint layers of different thicknesses attached to the surface of steel plates as the research object, and predicts and classifies the paint layer thickness based on the Transformer network algorithm. First, a steel plate paint layer thickness database is established, and after image preprocessing, it meets the requirements of Transformer network training and testing. Secondly, the Cross Former network (a multi-functional visual Transformer network model based on cross-scale attention) is selected to fine-tune the Cross Former network model. After network training and hyper-parameter optimization and adjustment, it can effectively predict the thickness of the paint layer. Again, by comparing the model training accuracy and test accuracy under different training hyperparameter settings, the optimal training hyperparameter values of the model are obtained, and the Cross Former network model under the optimal training hyperparameters is obtained. Finally, the polishing of the deep learning network prediction model is verified through experiments. The surface margin prediction theory was used and the paint layer thickness distribution map on the workpiece surface was drawn. The results showed that the real values in the experimental area were within the prediction range, which verified the effectiveness of the prediction model in actual application scenarios.

(2024) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Xingzuo Li, Chengzhi Su, Feng Jiang, Guang Zhao, and Yufei Mei "Research on feature surface thickness prediction based on transformer deep-learning-networks", Proc. SPIE 13184, Third International Conference on Electronic Information Engineering and Data Processing (EIEDP 2024), 1318466 (5 July 2024); https://doi.org/10.1117/12.3033021

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

;

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available

Members: $17.00

Non-members: $21.00 ADD TO CART

PROCEEDINGS
7 PAGES

DOWNLOAD PAPER SAVE TO MY LIBRARY

GET CITATION

RIGHTS & PERMISSIONS

Get copyright permission Get copyright permission on Copyright Marketplace

KEYWORDS

Education and training

Transformers

Deep learning

Data modeling

Head

Neural networks

Databases

Show All Keywords

Keywords/Phrases

Search In:

Publication Years