Rapid prediction method for elastane content in blended fabrics based on near-infrared hyperspectral imaging and...

The proportion of elastane fibers in blended fabrics generally ranges from 1% to 20%. In the process of recycling blended fabrics containing elastane fibers, the selection criteria of recycling processes for different elastane content ranges are as follows: when the content is 1%-5% (low content), the mechanical recycling method without separate separation is preferred; when the content is medium content, solvent separation and substrate recovery are adopted; when the content is high content, glycolysis is used. Inappropriate selection of recycling processes will lead to economic losses. Therefore, it is crucial to determine the elastane content prior to recycling. Current methods for detecting elastane content suffer from drawbacks such as long detection cycles and the requirement for sampling, making them unsuitable for large-scale rapid detection. To address this issue, this study proposes an elastane content prediction method based on hyperspectral imaging technology (spectral range: 1000-1700 nm) and neural network models. Experimental results demonstrate that the overall prediction accuracy of the proposed model reaches 0.895, verifying its effectiveness in elastane content prediction. This laboratory-scale study demonstrates the feasibility of quantitative elastane detection using hyperspectral imaging. It provides a proof-of-concept that lays a theoretical foundation for future industrial applications in the recycling and regeneration of elastane-containing blended fabrics, pending further large-scale validation.