Estimating Quality Components of Natural Fibers by Near-Infrared Reflectance1

Abstract

Near-infrared reflectance has been used to predict quality components of agricul tural products. Use of this technique for predicting cross-sectional area, specific surface (Arealometer A, mm2/mm 3), Micronaire reading, and causticaire maturity index of cotton fibers was explored. An instrument equipped with 1.53, 1.97, and 2.32 μm wavelength filters was calibrated to the properties of 20 reference samples. The properties of these samples were predicted and compared to values obtained by usual procedures. Correlation coefficients between predicted and observed values were r = 0.90 or greater for properties of the reference samples. Properties of three groups of samples that were not used in the calibration were predicted by the technique. Correlation coefficients between predicted values and those obtained by usual proce dures were r = 0.89 or greater. These data indicate that meaningful estimates of the four fiber properties can be obtained with near-infrared reflectance. Refinements in technique could improve the predictions. When performing near-infrared reflectance measurements, technician and instrument time per specimen is less than that with usual procedures except Micronaire reading. However, estimates of four properties of a specimen can be obtained in the same operation.