Automatic Weld Seam Inspection System Using Computer Vision and Artificial Intelligence

Abstract

The implementation of automated systems has become essential to increase the efficiency and reliability of industrial processes. In this context, the development of a weld seam quality inspection system is proposed, based on computer vision cameras and artificial intelligence algorithms, capable of identifying and classifying defects in welded joints in real time. The initiative aims to enable partial or total replacement of human inspection, enhancing the accuracy and standardization of evaluations. In doing so, it seeks to detect different types of faults, reduce inspection and rework costs, ensure consistent quality and traceability, increase operational efficiency, and create a structured database to support continuous improvement actions.