Quantitative Optical Identification of Fiber Composition in Mixed Textile Bales
Diffuse reflectance spectroscopy provides rapid non-destructive fiber identification in mixed bales when multivariate models correct for moisture and density.
Second Derivative
Mathematical Operation
NIR spectral pretreatment in textile testing often requires the elimination of broad baseline shifts caused by fabric texture and thickness. Applying a second derivative calculation to the raw absorbance data isolates the subtle changes in spectral curvature. This operation removes additive and linear baseline variations from the textile scan.
It enhances the peaks associated with specific fiber polymers and chemical finishes.
Baseline Correction
This mathematical transformation works by calculating the rate of change of the spectral slope at each wavelength. When the second derivative is computed, it converts broad bands into sharp, downward-pointing peaks that represent the true centers of absorption. This inversion facilitates the identification of highly specific chemical bonds.
It allows the software to focus on local spectral features rather than broad, non-specific shifts. This focus prevents physical scattering from yarn twist variations from skewing the chemical results.
Spectral Resolution
Overlapping peaks from fiber blends are separated by this derivative technique. In cotton-acrylic blends, the closely spaced nitrile and hydroxyl bands can merge into a single broad peak. Calculating the second derivative resolves this composite signal into its distinct component peaks.
Practical Implementation
Derivative processing can also increase high-frequency instrumental noise, which requires the simultaneous application of a smoothing filter. Technicians use the Savitzky-Golay algorithm to compute the derivative while suppressing electronic and detector noise. This balanced smoothing is necessary to maintain a high signal-to-noise ratio in the preprocessed spectra.
It ensures the resulting calibration model remains stable over time.