REAL-TIME WELDING TRAJECTORY IDENTIFICATION USING IMAGES IN ROBOTIC MANIPULATORS

Abstract

This paper presents a novel approach for real-time trajectory identification of welding paths in robotic manipulators using image processing techniques. The proposed method combines edge detection algorithms, deep learning models, and adaptive control strategies for accurate welding path tracking. Experimental results demonstrate a significant improvement in welding precision, path consistency, and processing speed. Key findings include a 25% increase in trajectory accuracy and a reduction in defect rates.