Physical uniformity of graded railcar and vessel shipments of Canada Western Red Spring wheat determined by digital image analysis

Abstract

The application of digital image analysis for objective measurement of grain uniformity in commercial samples of Canada Western Red Spring (CWRS) wheat was investigated. Official railcar-unloading (carlot) and vessel-loading (cargo) samples of three grades (Nos. 1, 2 and 3) of CWRS wheat were analyzed for 16 kernel-size, shape and brightness features that were automatically computed from digital images acquired in reflected light. Based on a subsample of 400 kernels for each graded sample, coefficient of variation (CV) data were used to evaluate the influence of grading and bulk handling on the uniformity of carlot and cargo shipments. A significant and progressive decrease in sample uniformity for No. 2 and No. 3 carlot samples, compared with No. 1, was observed for all computed kernel features. Averaged over all features, the decreases in uniformity in carlot samples were 13.4 and 13.7% with grade lowering from No. 1 to No. 2 and from No. 2 to No. 3, respectively. For cargo samples, an analogous effect was found as the grade dropped from No. 2 to No. 3 only. There was essentially no measurable difference in the level of uniformity, averaged over all kernel features, between No. 1 and No. 2 cargo samples, which were extremely uniform, as indicated by CV values of 1.27 and 1.25%, respectively. The effect of the bulk handling system was evidenced by a substantial improvement in uniformity of cargo samples compared with carlot samples. Averaged over all features, CV values for cargo samples were lower than CV values for corresponding carlots by 58, 63 and 64% for grade Nos. 1, 2 and 3, respectively. Key words: Wheat grading, digital image analysis, uniformity measurement