Defect classification is the process of identifying and categorizing faults or abnormalities in a product, component, material, or process according to predefined defect types. It often uses inspection data, images, sensor readings, or test results to distinguish acceptable items from specific defect categories, such as cracks, scratches, missing components, contamination, or dimensional errors.
In manufacturing, computer vision models can examine products on an inspection line and classify visible defects automatically, helping operators identify recurring quality problems. The classification result can then support sorting, rework, root-cause analysis, or process improvement. Depending on the application, defect classification may use machine learning, deep learning, image processing, anomaly detection, and automated optical inspection.
