Weld Seam Detection Based on the Genetic Algorithm

• Published in: Applied Mechanics and Materials Vol. 201-202; pp. 222 - 225
• Main Authors: Ding, Du Kun, Tan, Wei
• Format: Journal Article
• Language: English
• Published: Zurich Trans Tech Publications Ltd 01.10.2012
• Subjects: Genetic Algorithm (GA); Seam Weld; Weld
• ISBN: 9783037854822; 3037854820
• ISSN: 1660-9336; 1662-7482; 1662-7482
• DOI: 10.4028/www.scientific.net/AMM.201-202.222

A new weld seam detection based on the genetic algorithm has been presented. Firstly, the experimental platform has been setup, and the welding images can be captured by the CCD camera in real time. On this basis, the process region, which is in front of the weld pool, has been determined. Then the median filter has been applied to reduce the disturbance of the noise producing during the welding period. Moreover, the GA has been deeply researched and applied to the weld seam detection. In the end, some experiments have been caught out, and the results show that compared with the traditional Ostu process method, the weld seam images, which are processed by the genetic algorithm, are more clearer and have less noise, which can meet the satisfy of auto welding system.