Performance analysis of reduced-specification visible-near-infrared spectroscopy for municipal solid-waste classification

The use of near-infrared (NIR) spectroscopy for material classification in recycling facilities has become increasingly prominent due to its rapid, accurate and non-destructive capabilities. However, the high cost of commercial NIR spectrometers – driven largely by their wide spectral range and high resolution – limits widespread adoption, particularly in the smaller-scale decentralized waste segregation stations. This study investigates the feasibility of reduced-range and lower-resolution visible-near-infrared (VIS-NIR) spectra data for accurate waste classification. A dataset of 660 VIS-NIR spectra (500–2500 nm) was constructed by acquiring spectra from 11 classes of representative municipal solid waste samples. Using this dataset, we simulate spectral degradation by systematically reducing the wavelength range and resolution. A hybrid convolutional neural network - support vector machines (CNN-SVM) model is trained on both the full and reduced-specification spectra to assess the impact on classification performance. The results indicate that, for the dataset investigated, an optimally selected spectral window of approximately 150 nm was sufficient to achieve classification performance comparable to that obtained using the full spectral range (mean cross-validation accuracy of 98.43 ± 0.79%). Overall, these findings demonstrate the potential for reduced spectral requirements to enable simpler and potentially lower-cost NIR systems.