The objective measurement of alpha-amylase in wheat kernels using spectral imaging

When wheat kernels are wetted in the head prior to harvest, the germination processes are initiated. The symptoms range from no obvious visible signs of enzyme activation to gross kernel disfiguration. Alpha-amylase, a starch degrading enzyme is the most prevalent of the activated enzymes in the early stages of germination and may cause significant end-product quality loss. Current analytical techniques do not provide a rapid system for estimating individual kernel sprout damage. We have developed an objective approach using near-infrared spectra (1100-2400 nm) from a hyperspectral camera to predict α-amylase levels of individual kernels in two classes of Canadian wheat. Multivariate modeling gave, an R² of up to 0.69 for predicting individual kernel α-amylase levels. Using the hyperspectral data, a multispectral model predicted α-amylase activity levels of greater than 1 SKU unit/g with a better than 90% accuracy. At this level, there is no visible sign of kernel sprouting.