A Robust Real-Time Leather Defect Segmentation Using Yolo

Natural leather is a material made from animal skin and treated with chemical products to preserve it. It is used in the manufacture of clothing, bags, furniture, automotive material, and footwear, among others. Because of its importance in the industry, it is important to ensure its quality. Traditional inspection by a human expert is an expensive, time-consuming, and sometimes subjective process. Automatic inspection has become an essential part of any production system as it rejects nonconformities, ensures product quality, reduces operating costs, and shortens production cycle times. This paper presents a mathematical modeling approach using computer vision and artificial intelligence algorithms for the automatic inspection of natural materials - a leather case study. To achieve the best results in defect detection, the YOLO algorithm was chosen. More specifically, a comparison of the Small, Medium, Large, and Extra-Large models of YOLOv5 in leather defect detection was performed. The dataset MVTec Anomaly Detection was used to collect images of leather with and without defects. After training, the models were analyzed and compared based on some performance metrics. All models showed a great ability to detect defects in the dataset used.