Image of welding defects of energy storage welding machine

Why is it important to monitor welding defects?

Welding quality plays a vital role in the durability and effectiveness of welding structures. Therefore, it is essential to monitor welding defects to ensure welds quality. Manual inspection, analysis and evaluation of welding defect images is difficult due to the non-uniformity in their shape, position, and size.

How are welding defects identified?

Welding defects can be identified using various inspection methods, which can be broadly categorized into non-destructive testing (NDT) and destructive testing. 1. Visual Inspection The simplest and most cost-effective NDT method, where a trained inspector examines the weld visually for surface defects such as cracks, porosity, and misalignment.

Why do we need deep learning to identify welding defects?

Manual inspection, analysis and evaluation of welding defect images is difficult due to the non-uniformity in their shape, position, and size. Hence the use of deep learning techniques to identify welding defects is more accurate and reliable due to the adequate training data samples, which helps to identify welding defects with greater accuracy.

What is a weld defect?

This defect typically manifests as protrusions, uneven or overly pronounced weld beads that extend beyond acceptable dimensions, creating surface irregularities such as bumps or ridges.

What affects the quality of the welding process?

Minor deviations in the welding process can result in serious defects, like collapse, cracks, porosity, burn, welding hole, etc, thus affecting the quality of the welding process , . In addition, welding quality is also affected by the types of welding materials used and the types of welded joints , .

How many defect classes are there in a laser welding defect dataset?

The main contributions of this paper are: A novel laser welding defect dataset is developed that includes seven defect classes and one normal class. The up-sampling technique is used to overcome the imbalanceness in the dataset according to the majority class labels.