Shelton Vision’s systems use sophisticated algorithms to automatically detect defects in fabrics. These algorithms analyse large datasets of fabric images, enabling them to identify even subtle defects that might be missed by human inspectors. The system can recognise various types of defects, such as broken yarns, stains, or misprints, with high precision.
The AUTOTRAIN feature utilises advanced analysis to evaluate each fabric type automatically. This functionality enables the system to create and store optimal inspection settings for each fabric style, ensuring consistent and accurate defect detection across different fabric types, including dark and heavily textured materials. The system adapts to new fabric styles without requiring manual adjustments.
The CBIGS (Classification and Grading) engine employs AI to classify and grade defects in real-time. This AI-driven engine can filter out non-serious defects, prioritising significant issues that require attention. By doing so, it saves considerable manual labour and enhances the accuracy of defect reporting. The system can classify defects by type and severity, providing a detailed defect map.
The WebVIEWER software integrates advanced algorithms to control the rolling, stopping and display of instructions on re-roll machines. This ensures precise positioning for cutting and other operations, maximising throughput and minimising errors. The software interprets the optimised cut plan and directs machine actions accurately.
Advanced algorithms generate optimised cutting plans based on the defect map reviewed and adjusted in the previous steps. These algorithms consider stored rules related to customer and product requirements to create a cut plan that maximises yield and minimises waste. If the initial plan doesn’t meet yield requirements, the system can quickly reprocess the data to generate a new, improved plan.
Shelton Vision’s systems continuously learn and improve over time. As more data is processed, the algorithms become more accurate and efficient at detecting and classifying defects. This ongoing learning process ensures that the systems stay up-to-date with new fabric types and evolving quality standards.
