Defect detection of catenary cotter pin based on three-level cascade architecture and attention mechanism

Cheng Yue; Yaqi Hu

doi:10.1117/12.2672144

21 March 2023 Defect detection of catenary cotter pin based on three-level cascade architecture and attention mechanism

Cheng Yue, Yaqi Hu

Proceedings Volume 12609, International Conference on Computer Application and Information Security (ICCAIS 2022); 126090P (2023) https://doi.org/10.1117/12.2672144
Event: International Conference on Computer Application and Information Security (ICCAIS 2022), 2022, ONLINE, ONLINE

Abstract

For the problems of few samples, serious unbalanced categories and small defect scale of catenary cotter pin, a detection method based on three-level cascade architecture and attention mechanism is proposed. First of all, to alleviate the problem of inaccurate location caused by the small size of the cotter pin, the deep residual location network was constructed, and the atrous convolution layers of different sizes were introduced, which not only avoided the loss of internal data structure, but also preserved the hierarchical context feature. Secondly, a lightweight generative adversarial network that is sensitive to the difference of image local features is constructed to generate defect samples. Therefore, the problem of over-fitting and poor generalization performance of the detection network due to few samples and unbalanced categories is alleviated. Finally, in order to further alleviate the problem of small defect scale, the attention mechanism is adopted to learn different weights from channel domain, so as to obtain more important feature information and improve the accuracy of the detection network. Comparing the proposed method and its variant on dataset, the results show that the three-level cascade architecture proposed has certain advantages, and can alleviate the problems of few samples, unbalanced categories and small target scale.

Citation Download Citation

Cheng Yue and Yaqi Hu "Defect detection of catenary cotter pin based on three-level cascade architecture and attention mechanism", Proc. SPIE 12609, International Conference on Computer Application and Information Security (ICCAIS 2022), 126090P (21 March 2023); https://doi.org/10.1117/12.2672144

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
8 PAGES

DOWNLOAD PAPER SAVE TO MY LIBRARY

GET CITATION

RIGHTS & PERMISSIONS

Get copyright permission Get copyright permission on Copyright Marketplace

KEYWORDS

Object detection

Defect detection

Feature extraction

Convolution

Education and training

Detection and tracking algorithms

Image enhancement

Show All Keywords

Keywords/Phrases

Search In:

Publication Years