The objectives of this project were to compare results from Uster Technologies IntelliGin system, an in-line leaf, colour and moisture monitor for cotton ginning, with leaf and colour grade results from the standard High Volume Instrument (HVI), also built by Uster. We studied the statistical relationships between the leaf and colour grades measured by HVI instrument with those measured by the IntelliGin system. The datasets with corresponding HVI and IntelliGin results were supplied from three gins representing three growing regions. From each region between 65 and 67 sets of HVI data, representing about 200 bales after module averaging, were gathered. While this is only a small snap shot of Australian cotton production, the dataset reflects standard leaf and colour grade values found in Australian cotton. It is noted the subsequent relationships measured would have been improved with a larger set of samples containing a wider range of USDA leaf and grade classifications.

The main findings from this study are:

• Leaf (particle) count was found to be a better regressor for leaf grade than leaf (particle) area, although the USDA classification standard stipulates that leaf area is correlated to classer grade. The implication of this is that removing leaf from cotton by mechanical actions during ginning may not be the preferred option. Leaf count may be decreased by reducing mechanical actions during processing, leading to smaller number of larger leaf particles rather than a large numbers of pepper-sized particles. It is well documented that reduced mechanical action in the gin reduces the amount of damage to cotton and is generally beneficial to cotton quality in terms of fibre length and neps. In addition, fewer larger leaf particles are easier to remove than many pepper-sized leaf particles during cotton spinning. We recommend further study on this strategy.

• All bales examined were assigned the best pigmentation grade “white”, i.e. no samples were classed as “spotted” or worse. The +b values of all bales was consistently low (very few bales had +b values above 8) in terms of yellowness. The low +b values had an adverse effect on the USDA classing grade that was applied to the cotton. This matters particularly in the movement of the grade from Middling to Strict Low Middling, which carries a large discount on the cotton. In other words, a superior attribute of these cottons was penalised according to the current colour standard designed for American Upland Cotton. We understand this issue is being looked at currently by the Cotton Classers Association of Australia. This study provides more evidence for an opportunity to market Australian cotton as being “super white (very low +b value).

• Although the overall average values of leaf and colour grades from the two systems were quite consistent the values assigned to individual bales differed widely. Overall, 57% of the bales were assigned to the same leaf grade by the HVI and the IntelliGin; 40% of the bales were assigned the same brightness grade by the two systems.

• Very large differences in average values were found in the results measured by IntelliGin and by the HVI on bales, for the additional batch of grab samples examined. These differences could be attributable to the sampling rate of the manual grab versus the IntelliGin grab and baling rate, and the compression history of the samples, as well as the instrument differences already observed in the comparison between paired HVI and IntelliGin values from the three regions. However, without more consistent data these conclusions remain speculative.