Surface defect detection projects: scratches, dents, anomalous textures, coating defects on metal, glass, plastic, and laminated substrates. Deep learning handles the part-to-part variability that defeats rule-based vision.