The statistical technique of principal components is used to analyze two sets of near-infrared spectra, wheat flour samples for which % moisture and % protein values are included, and milled barley samples for which hot water extract values are included. The methodology and interpretation of this technique are described within the context of NIR data, and its advantages both in providing insight into the variation of the spectra, and as a method of avoiding the problems caused by highly correlated reflectance energy values in the derivation of calibration equations, are highlighted. In each set of samples the first principal component accounts for the vast majority of the variation. These components also have an almost identical shape, which is interpreted as reflecting particle size. The second wheat component and the third barley component are also almost identical, with a shape very similar to that of the spectrum of water. Both fourth components share peaks at points in the spectrum which are used by fixed-filter instruments to measure protein in cereals.

No abstract

The hardness of samples of 41 barley cultivars grown in Australia was determined by measuring milling energy. The milling energy was negatively correlated (r= -0.49) with the starch and positively correlated (r=0.50) with the p-glucan content. The correlation with protein content was not significant. This suggests that a low starch content and a high p-glucan content may contribute to hardness but other factors may probably also be involved. The extent of modification of these barley samples measured by Calcofluor staining after steeping and 48 h of germination was correlated with the grain hardness (r= -0.56). Factors contributing to grain hardness may limit the rate of endosperm modification during malting, indicating the value of selecting softer cultivars for malting.

Cereal Chem. 78(5):572-577Three problems need to be addressed in networks of Infratec Grain Analysers: 1) the networks are not interconnected, 2) the partial least squares (PLS) calibrations used so far have to be individually adjusted for bias when transferred to the slave instruments, and 3) the calibrations are not entirely stable over time. Nonlinear artificial neural network (ANN) calibrations based on a large common European data set (≈4,000 samples in the training sets and ≈1,000 samples in the stop sets) were introduced to overcome these constraints. The performance of these ANN calibrations was compared with Danish PLS models for protein and moisture in cereals during the 1998 harvest in Denmark, and subsequently with PLS models based on the same European data set. ANN models were more accurate than PLS and, unlike PLS, were linear and transferable up to 25% moisture. It is suggested that the improved performance of the ANN models is attributable to the modeling technique rather than the size and nature of the European data set. In most cases, ANN models could be applied directly and without bias adjustment to slave instruments. The ANN models were also more stable, they required fewer bias adjustments or remodeling over time compared with Danish PLS models. ANN calibrations using shared data have been adopted for commercial use in several European countries and work is in progress to develop global ANN models for determination of protein in wheat and barley.

Near infrared reflectance spectroscopy in the very near infrared region (700 nm to 1100 nm) has been investigated for the detection of grain weevil larvae and pupae inside single wheat kernels. Using a total of 80 samples, simple, two-wavelength classification models have been identified, based on either log 1/R (982 nm)-log 1/R (1014 nm) or log 1/R (972 nm)-log 1/R (1032 nm). Both models correctly classified over 96% of samples as uninfested or infested. Detection performance equalled that obtained using the full-spectrum approach of principal components analysis. In a separate experiment, repeatedly scanning samples over time demonstrated detection of younger larvae as well as later developmental stages. This experiment confirmed that the observed spectral differences arise from the actions of the developing insect, rather than from any feature specific to kernels selected by adult females for egg-laying. The origins of the spectral differences are almost certainly decreasing grain starch, for log 1/R (982 nm)-log 1/R (1014 nm), or increasing grain moisture, for log 1/R (972 nm)-log 1/R (1032 nm), with infestation. These results indicate that the future incorporation of the wavelength pair, 982 nm and 1014 nm, as camera lens filters in a very near infrared imaging system, could lead to an inexpensive, rapid and reliable machine-vision method for detecting internal insects in grain.