AI-Powered Textile Classification for Automated Sorting
Specim has released their ground-breaking RETEX system, combining powerful hyperspectral technology with AI capabilities for highly-accurate identification of textiles, including complex blends and impurities, enabling real-time automated material separation.
Specim RETEX delivers a streamlined material processing solution, designed for ease of integration into existing processes without requiring expensive redesign. It offers very high processing speeds, enabling high throughput of up to 5 garments per second while maintaining unmatchable classification accuracy.
Specim RETEX Key Capabilities:
- Pure cotton identification with 99% accuracy
- Pure polyester identification with 98% accuracy
- Polyamide identification with 99% accuracy
- Identification of cotton and polyester percentage in polycotton blend with 5% error margin
- Viscose, wool and acrylic identification
- Elastane detection, including in previously unseen blends
- Detection of impurities and colour recognition
- Real-time processing for production environments
Flexible Solution Options
Specim RETEX is available in four different variants to support different use cases:
- Specim RETEX Core - AI classification engine for integration into custom systems.
- Specim RETEX Lab - Hyperspectral textile analysis for R&D and validation.
- Specim RETEX Modules - Hyperspectral and AI components for textile recognition, designed for OEMs, integrators and research.
- Specim RETEX System - Complete industrial solution for automated textile processing and material separation.
The modular architecture enables deployment from lab environments to full-scale industrial systems.
“Hyperspectral imaging allows us to identify textile materials with a level of accuracy not achievable with conventional vision systems,” said Sami Toljamo, CCO of Specim, Spectral Imaging Ltd. “With Specim RETEX, we combine this capability with AI-driven analysis to enable reliable material recognition and scalable textile processing.”
