Weld defect identification through 3D scanning and error distribution modelling supported by point-feature histogram

Hegedűs-Kuti, János (2026) Weld defect identification through 3D scanning and error distribution modelling supported by point-feature histogram. GÉP, 77 (1). pp. 10-14. ISSN 0016-8572

Preview

Text gep2026-1-2.pdf - Published Version Download (338kB) | Preview

Abstract

The inspection of welded structures requires specialattention, as the quality of the product is influenced bynumerous factors. Determining the suitability of weldsis a complex task. The aim of this work is to develop amethod that effectively supports this evaluation process.The study presents a modern evaluation procedurebased on PFH data and its practical application in theinspection of weld quality. Reference and defective pointclouds were created using image processing and CADmodeling software, and then aligned using the ICPalgorithm. This serves as the basis for generating datasets representing model errors. The results show thatthe distributions of the variables obtained in this waydescribe the nature of typical welding defects well.Automated evaluation of these distributions can reducethe turnaround time of inspections, which contributes toincreasing the efficiency of welding processes.

Actions (login required)

Edit Item