Development of a Practical Visualization System for Gas Metal Arc Welding Skill Training Using Image Processing Techniques

Le résumé fourni par la source

Observation of welding features is important for GMAW training and instruction because the welding arc, molten pool, filler wire, and groove can be difficult to distinguish during welding. In this study, a compact, low-cost, and practical visualization system was developed to support gas metal arc welding (GMAW) skill training from both the welder’s and instructor’s perspectives. The system consists of a welder-side unit and an instructor-side unit and uses a commercial camera, optical filters, a wide-angle lens, and a compact computer. Welding images were acquired under actual GMAW conditions, and the effects of optical filter selection, exposure time, tone mapping, and trimming methods were investigated. A 600 nm long-pass filter and an exposure time of 20,000 μs provided a suitable balance between arc-light suppression, brightness stability, and image clarity. Gamma correction improved the visibility of key regions, including the molten pool, arc, torch, groove, and wire. In addition, low-pass-filtered centroid tracking enabled stable trimming of the weld region from wide-angle images. The developed system achieved real-time display and recording of standardized welding images, demonstrating its potential to support GMAW training through improved image visibility, real-time monitoring, and standardized image recording, while also providing visual data for post-weld review and future skill-assessment applications.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.