An application of computer vision-based material classification and fragility risk assessment

This independent study explores the application of computer vision-based material classification with fragility risk assessment to enhance process efficiency in logistics and supply chain management. The proposed system leverages advanced computer vision model, YOLOv8-seg to accurately segment between three material types: ceramic, plastic, and metal. In addition to segmentation, the system assesses fragility using product shape, enabling proactive handling strategies to minimize damage during transportation and storage. By model material classification and fragility evaluation, the system reduces manual effort, minimizes errors, accelerates product sorting, and optimizes packaging decisions.