In this experiment, poplar fibers containing 0%, 2%, 4%, 6%, 8%, 10%, 12%, 14%, 16%, 18%, 20%, 25% and 30% of urea-formaldehyde resin were prepared. A model for the detection of urea-formaldehyde resin content in poplar fibers was established by the hyperspectral near-infrared imaging system combined with relevant algorithms. The spectral images of poplar fibers containing different contents of urea-formaldehyde (UF) resin were measured separately using hyperspectral imager. The results of four preprocessing methods, namely mean centering (MC), multiple scattering correction (MSC), standard normal variables (SNV) and first-order derivative (1-Der) were analyzed, and the optimal preprocessing method was selected as SNV. The band combinations with the highest correlation with the urea-formaldehyde resin content were compared and analyzed with the full-band model to establish the partial least squares regression (PLSR) model. The experimental results show that the hyperspectral imaging system combined with the corresponding algorithm can achieve rapid detection of UF resin content in poplar fibers, and the results of this study provide technical support and theoretical reference for determination of resin content in ultra-thin fiberboard production. The method is an innovative model for the determination of UF resin in wood fibers.
Read full abstract