Research on Rapid Identification Technology of Microfiber Evidence in Criminal Cases Based on Raman Spectroscopy and Deep Learning

Rapid and accurate identification of trace microfiber evidence in criminal investigations is essential for forensic analysis and relies on precise characterization of textile materials. As an important electromagnetic wave–matter interaction technique, Raman spectroscopy provides non-destructive optical sensing capabilities for revealing the molecular composition and structural characteristics of fiber samples. This study integrates Raman spectroscopy with a deep learning framework to establish an intelligent identification system for microfiber evidence. A total of 8,500 high-quality Raman spectra covering 15 common and blended fiber types were collected, and a preprocessing workflow including baseline correction, smoothing, denoising, and normalization was developed. A one-dimensional convolutional neural network was then constructed for spectral feature extraction and classification. Experimental results demonstrate that the proposed model achieves an overall recognition accuracy of 98.7% on an independent test set. The results confirm that the combined approach effectively captures the chemical signatures of complex textile fibers and enables reliable identification of trace evidence, providing a robust technical solution for forensic analysis and electromagnetic spectroscopy-based intelligent material recognition.