Textile Analysis for Recycling Automation using Transfer Learning and Zero-Shot Foundation Models

2025-06-06

AI Summary

To address the challenge of automating textile recycling, this paper proposes a dual-task vision analysis framework based on RGB images: (1) classification of four common textile materials and (2) pixel-level segmentation of non-fabric foreign objects (e.g., buttons, zippers). Methodologically, we innovatively cascade the zero-shot open-vocabulary detector Grounding DINO with the Segment Anything Model (SAM) to achieve annotation-free foreign-object segmentation; concurrently, we employ transfer learning with EfficientNet-B0 to enhance material classification performance. Our approach eliminates reliance on expensive specialized sensors and labor-intensive pixel-level annotations. Evaluated on real-world recycling scene data, the framework achieves 81.25% accuracy in material classification and an mIoU of 0.90 for foreign-object segmentation. These results demonstrate both the effectiveness and state-of-the-art capability of low-cost RGB vision in pre-processing for textile recycling.

Technology Category

• Computer Vision: Segmentation
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Natural Language Processing: Sentence-level Semantics, Textual Inference, etc.

Application Category

• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Web Mining and Content Analysis: Normalization, clustering, classification, and summarization of Web text
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Automated sorting is crucial for improving the efficiency and scalability of textile recycling, but accurately identifying material composition and detecting contaminants from sensor data remains challenging. This paper investigates the use of standard RGB imagery, a cost-effective sensing modality, for key pre-processing tasks in an automated system. We present computer vision components designed for a conveyor belt setup to perform (a) classification of four common textile types and (b) segmentation of non-textile features such as buttons and zippers. For classification, several pre-trained architectures were evaluated using transfer learning and cross-validation, with EfficientNetB0 achieving the best performance on a held-out test set with 81.25% accuracy. For feature segmentation, a zero-shot approach combining the Grounding DINO open-vocabulary detector with the Segment Anything Model (SAM) was employed, demonstrating excellent performance with a mIoU of 0.90 for the generated masks against ground truth. This study demonstrates the feasibility of using RGB images coupled with modern deep learning techniques, including transfer learning for classification and foundation models for zero-shot segmentation, to enable essential analysis steps for automated textile recycling pipelines.

Problem

• Automated textile sorting using RGB imagery for recycling efficiency
• Classification of textile types via transfer learning techniques
• Segmentation of non-textile contaminants with zero-shot foundation models

Innovation

• Uses RGB imagery for cost-effective textile analysis
• Employs transfer learning with EfficientNetB0 for classification
• Combines Grounding DINO and SAM for zero-shot segmentation