Research on Weld Strengthening Based on Machine Vision Technology

Abstract

Welding, as an important process in industrial manufacturing, is often accompanied by various forms of light, heat, and electromagnetic radiation, which can pose significant harm to the human body. The modern industry has effectively mitigated direct harm to humans from welding by introducing robots and embracing intelligent manufacturing, thereby enhancing the quality of welding. Traditional welding quality inspection relies on human eyes, requiring extensive experience and resulting in low detection efficiency. Due to the complexity of the welding process and the randomness of process interference, false detection and missed detection are inevitable during testing. With the rapid development of machine vision, characterized by precision and high intelligence, it has found wide applications in industrial measurement, product testing, and identification. However, machine vision methods impose high requirements on image quality. Utilizing MATLAB to build the algorithm platform, the methods of gray scale conversion, image filtering, and histogram equalization have been respectively adopted to enhance weld image processing. The experimental results demonstrate that histogram equalization holds practical engineering significance for the research objects selected in this paper, effectively improving the image quality of the inspected weld.