Abstract

Quality inspection is a fundamental pillar of textile manufacturing, as garment defects directly affect customer satisfaction, production efficiency, and overall brand reputation. In this context, automated inspection systems have become essential for ensuring consistent product quality and reducing reliance on manual inspection, which is often labor-intensive, inconsistent, and susceptible to human error. With the emergence of industry 4.0 and the increasing adoption of automation and smart manufacturing technologies in the textile sector, the demand for intelligent and automated quality inspection systems has significantly increased. Recent advances in deep learning and computer vision have opened new opportunities for precise and real-time identification of sewing defects. This study proposes a YOLOv11-based framework for detecting critical defects such as crooked seams and skipped stitches, aiming to enhance accuracy, speed, and reliability in garment inspection. The experimental results demonstrate the potential of the proposed method to significantly improve quality assurance processes within modern apparel manufacturing environments.

IPC Classification

Keywords