Canadian Grain Commission
Symbol of the Government of Canada

Measuring barley kernel colour and size to predict end use malt quality

Results and discussions - Barley kernel colour and malt quality

The ability of objective colour measurement to predict malt quality was compared to the prediction by degree of soundness in the barley grades. In the 68 samples tested, Special Select barley samples showed significantly better barley kernel colour (Table 3) than the other two grades, while colour of Select barley was significantly better than that of Standard Select. This was expected given that colour-related factors, kernel staining and discolouration, were the major degrading factor for these samples. The colour value from imaging correlated significantly with Inspector's colour figure (r=-0.87, p<.001), the negative coefficient a result of good colour being represented by high numbers with imaging and low numbers with the Inspector's system.

Table 3. Average of imaging and visual colour values for Canadian Grain Commission malting grades
CGC Malting Barley Grade Image Colour (RGB green) Inspector Colour
a, b, c row means with different superscripts, differ significantly (p<0.01) based on Duncan's Multiple Range Test
Special Select 171 a 1.0 c
Select 166 b 1.6 b
Standard Select 158 c 3.7 a

Correlations between malt quality and grades or colour measurement by imaging suggested that additional information was provided by the machine-based colour results (Table 4). Grade, and both image and inspector's colour measurements, had similar significant correlations with steep-out moistures, soluble protein and alpha-amylase, although, correlation coefficients for the image results were always slightly higher. Image colour also showed significantly higher correlations, than the other two methods, for 1000 kernel weight, germinative index, malt extract and diastatic power.

In summary, the three methods for quantifying appearance all showed ability to predict end use quality. Results from image analysis, though, were significantly better for several analyses and most likely more objective compared to the visual methods that could show significant variability among inspectors. The present study suggested further testing of the imaging system using samples with additional impediments to appearance and comparisons with results from several different inspectors.

Table 4. Correlation coefficients between indices of kernel colour and malt quality parameters (n=68)
  Grade Image Colour ¹ Inspector's Colour
¹ RGB green
*, **, *** coefficients significant at: p<0.05, p<0.01 and p<0.001 levels, respectively
Image colour -0.75*** 1.00
Inspector's colour 0.74*** -0.90*** 1.00
1000 kernel weight 0.21 -0.55*** 0.45***
Barley protein 0.09 0.00 0.01
Germination energy (4 mL) 0.11 0.08 -0.06
Germinative index (4 mL) 0.21 -0.32** 0.16
Steep-out moisture 0.51*** -0.57*** 0.45***
Malt extract -0.02 -0.25* 0.20
Soluble protein 0.57*** -0.63*** 0.52***
ß-Glucan 0.05 -0.14 0.15
Diastatic power 0.28* -0.44*** 0.28*
α-Amylase 0.41*** -0.48*** 0.33**
Wort colour 0.12 -0.08 0.15
Friability -0.31* 0.50*** -0.42***