Submitted to: Applied Engineering in Agriculture
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 6/29/2011
Publication Date: 10/11/2011
Citation: Liu, Y., Gamble, G.R., Thibodeaux, D.P. 2011. Potential of near infrared spectroscopy in the prediction of cotton fiber strength indices. Applied Engineering in Agriculture. 27(5): 839-843.
Interpretive Summary: Cotton fiber strength is one of important quality characteristics and is related with the manufacturing of quality goods for consumers. Currently, two types of instrumentals have been developed to assess the cotton fiber strength, namely, automation oriented HVI and laboratory based Stelometer device. Each of these instruments has its unique advantages, and apparently, correlation between the two strength readings was relatively low. Meanwhile, HVI strength based near infrared (NIR) models were observed to lack the power for practical applications. Although both HVI and Stelometer strength readings have been adapted in numerous publications, none has attempted to unravel the relationship between the two. In this study, both HVI and Stelometer strength readings were corrected by respective micronaire values, and then the effect of modified HVI strength values on NIR model performance were examined through partial least squares (PLS) regression. 2 modified strength readings were observed to have a much improved relationship, and also corrected HVI strength was found to have a better correlation with NIR spectra and resultant model was acceptable for quantitative determination of cotton strength index. The outcome provides cotton fiber / textile engineers, researchers and regulators a new sight in applying both HVI and visible/NIR spectroscopy for rapid and routine determination of cotton strength qualities.
Technical Abstract: Despite relatively low correlation between 2 cotton strength readings from the automation oriented HVI and laboratory based Stelometer device, the present study demonstrates the consistence of cotton fiber strength measurements between the two methods if the strength readings were modified by cotton fiber micronaire property. The corrected HVI strength indices were found to have better NIR spectral response than uncorrected strength values, through PLS regression. Especially, the use of quotient form resulted in a more improved model performance than that of product one, suggesting the potential of NIR technique in the prediction of cotton fiber strength index for quality control application. On the basis of small sample numbers from 2 NIR instruments, this study indicates a “proof-of-concept” and a promising NIR model development.