Development of a convolutional autoencoder using deep neuronal networks for defect detection and generating ideal references for cutting edges

Abdullah Karatas; Dorothea Kölsch; Samuel Schmidt; Matthias Eifler; Jörg Seewig

doi:10.1117/12.2525882

21 June 2019 Development of a convolutional autoencoder using deep neuronal networks for defect detection and generating ideal references for cutting edges

Abdullah Karatas, Dorothea Kölsch, Samuel Schmidt, Matthias Eifler, Jörg Seewig

Author Affiliations +

Proceedings Volume 11056, Optical Measurement Systems for Industrial Inspection XI; 1105623 (2019) https://doi.org/10.1117/12.2525882
Event: SPIE Optical Metrology, 2019, Munich, Germany

Abstract

Cutting edges are of great importance in industry and especially in mechanical engineering. However, like other components, they wear out over time. The contour and in particular the cutting edge itself can be damaged over time or by other occurrences and be defective. If the traces of use or defects are small, they can be corrected by reworking. This means that the cutting edge can still be used by post-processing. To achieve this, it is necessary to measure the cutting edge. Subsequently, the error must be evaluated. This error should indicate whether and how far the cutting edge must be reworked. In order to carry out such an evaluation, ideal references of the cutting edge are necessary. If an ideal geometry of the cutting edge is available as a computer-aided design model, the evaluation is trivial. However, this only exists in very rare cases. Often the reference geometry must be formed on the basis of one measurement. This paper presents a possibility of reconstructing cutting edges and therefore a rating of this cutting edge. The reconstruction is based on neuronal networks, more precisely by convolutional neuronal networks.

Conference Presentation

Citation Download Citation

Abdullah Karatas, Dorothea Kölsch, Samuel Schmidt, Matthias Eifler, and Jörg Seewig "Development of a convolutional autoencoder using deep neuronal networks for defect detection and generating ideal references for cutting edges", Proc. SPIE 11056, Optical Measurement Systems for Industrial Inspection XI, 1105623 (21 June 2019); https://doi.org/10.1117/12.2525882

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