Click on “Download PDF” for the PDF version or on the title for the HTML version.
If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.
Visible and Near-Infrared Instruments for Detection and Quantification of Individual Sprouted Wheat Kernels
Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan www.asabe.org
Citation: Transactions of the ASABE. 59(6): 1517-1527. (doi: 10.13031/trans.59.11566) @2016
Authors: Paul R. Armstrong, Elizabeth B. Maghirang, Kevin F. Yaptenco, Tom C. Pearson
Keywords: Falling number, Near-infrared, Sorting, Sprouting damage, Wheat.
Abstract. Pre-harvest sprouting of wheat kernels within the grain head presents serious problems as it can greatly affect end-use quality. Functional properties of wheat flour made from sprouted wheat result in poor dough and bread-making quality. This research examined the ability of two instruments, a single-kernel near-infrared (SKNIR) spectroscopy instrument and a silicon light-emitting diode (SiLED) sorter, to estimate the level of sprouting in single kernels of wheat in order to provide breeders with a quick method to evaluate sprouting resistance. The SKNIR uses near-infrared reflectance (NIR) spectra spanning 990 to 1700 nm, while the SiLED uses a silicon sensor to measure visible reflected light obtained from a kernel using several LEDs at discrete wavelengths. Multiple varieties of hard white (HW) and soft red winter (SRW) wheat were conditioned to nine sprouting levels, S1 (untreated) to S9 (highly sprouted), using a multi-step soaking protocol, with falling number (FN) determined from subsamples of each sprouting level. Partial least squares model predictions of FN based on single-kernel NIR spectra yielded R2 values of 0.59 to 0.72. This improved when the spectral averages from 30 kernels were used to predict FN, resulting in R2 values of 0.78 to 0.95. Sprouting that was separated into three levels (sound, intermediate, and severely sprouted) could be classified reasonably well using the NIR instrument. The SiLED sorter was used to sort samples at each of the nine sprouting levels into two classes (sound and unsound kernels), with sound kernels re-sorted twice. Sprouted kernels were then visually identified from the three sound streams and quantified by weight. Linear models developed based on these weights were used to predict FN and sprouted kernels. Models were marginal in most cases, with R2 values ranging from 0.59 to 0.87. Overall, both instruments can identify samples with minimal or no sprouting and those with more severe sprouting; kernels with intermediate levels of sprouting were difficult to quantify.
(Download PDF) (Export to EndNotes)