Automatic inspection of surface defects in synthetic leather using optical techniques combined with image processing

Abstract Artificial leather (synthetic leather) is one of the most widely used materials in the textile industry. During the production process, raw material sheets undergo several stages such as color formulation, polymer coating, thermal processing, and lamination. These steps can lead to surface defects such as scratches, wrinkles caused by pressure, and stains from dyeing. Rapid and accurate detection of these defects in a stable and continuous manner remains a major challenge for product quality inspection systems. Although manual inspection by human eyes can detect defects effectively, it lacks the durability and repeatability of automated systems. However, current automated inspection systems still face difficulties when operating over large surface areas and identifying a wide variety of defect shapes (scratches, peeling, wrinkling) and color variations (stains, discoloration), which significantly reduces their accuracy. In this study, we propose an optical system that utilizes dome illumination, designed to operate effectively over a wide inspection area and generate high-contrast images that highlight defect locations on the leather surface. The resulting images are then processed using image processing techniques to detect and localize defects, supporting decision-making for accepting or rejecting products in real-time production lines.