Related Experiment Video
Updated: Sep 22, 2025
06:21
A Simple and Scalable Fabrication Method for Organic Electronic Devices on Textiles
Published on: March 13, 2017
10.5K
Research on Online Rapid Sorting Method of Waste Textiles Based on Near-Infrared Spectroscopy and Generative
Jinquan Hu1,2, Huihua Yang1,3, Guoliang Zhao4
1School of Artificial Intelligence, Beijing University of Posts and Telecommunications, 10 Xitucheng Road, Haidian District, Beijing, China.
Computational Intelligence and Neuroscience
|May 24, 2022
Summary
This study introduces a new method for rapid online sorting of waste textiles using a generative adversarial network and a classification model. The approach effectively identifies pure textile spectra, offering high speed and robustness for recycling applications.
Area of Science:
- Materials Science
- Computer Science
- Environmental Science
Background:
- Waste textile sorting is crucial for effective recycling and resource management.
- Current methods face challenges in accurately identifying blended textile compositions.
- High-content blending data is essential for developing robust sorting algorithms.
Purpose of the Study:
- To develop an automated system for rapid online sorting of waste textiles.
- To generate effective high-content blending spectral data using a generative adversarial network (GAN).
- To construct a robust classification model for identifying pure textile spectra within mixed waste.
Main Methods:
- Utilized a generative adversarial network (GAN) to mine combination relationships within blending spectra.
- Developed a classification model by integrating Generative Adversarial Networks (BEGAN), Radial Basis Function (RBF), and Support Vector Machines (SVM).
- Compensated for negative sample imbalance within the dataset to improve model accuracy.
Main Results:
- The developed model effectively extracts spectra of pure textile samples from complex blends.
- Experimental results demonstrate high robustness and speed of the classification model.
- The model achieves performance comparable to leading global technologies in textile sorting.
Conclusions:
- The proposed BEGAN-RBF-SVM model offers a highly effective solution for online rapid sorting of waste textiles.
- The method demonstrates significant potential for improving textile recycling efficiency and market applicability.
- Accurate spectral extraction and classification are key to advancing waste textile management.
