Use Of Artificial Neural Networks For The Correction Of Faults On The Detected Weld Seam Image
Loading...
21.pdf (2.15 MB)
Authors
Date
2017
Journal Title
Journal ISSN
Volume Title
Publisher
BGJ Consulting
Abstract
"Recently, the use of industrial robots has become widespread and the number of researches on this area has also increased considerably. Particularly the work done on robotic welding process is striking. Much of this work is about improving welding quality and automatically determining the welding path. Generally, 2D and 3D image processing methods have been used to determine the weld seam with various algorithms. Previous studies show that the biggest problem in this methods applied is the clear separation of the weld seam from the image. In this experimental study, artificial neural networks are used as a solution for this common problem. The most accurate result are obtained with the applied algorithms. It can be seen that the obtained results show the weld seam can be detected clearly using artificial neural networks."
Description
Keywords
Citation
Yildirim Ş., Ulu B., Use Of Artificial Neural Networks For The Correction Of Faults On The Detected Weld Seam Image, [in:] Journal of Current Construction Issues. CIVIL ENGINEERING PRESENT PROBLEMS, INNOVATIVE SOLUTIONS - Optimization in Business and Engineering, ed. Jarosław Górecki, BGJ Consulting, 2017