Enhancing wheat quality through color sorting: a novel approach for classifying kernels based on vitreousness

Abstract

Introduction: Wheat is a major food crop used in producing bread, noodles, and cookies. Kernel vitreousness, closely related to protein content, is key to determining wheat's processing purpose. Traditionally, vitreousness is visually assessed, but studies on classifying vitreous and starchy kernels to improve quality are limited.

Methods: This study expands the use of a commercial color sorter to classify kernel vitreousness by G value, distinguishing vitreous from starchy kernels.

Results and discussion: The system improved protein content and bread-making quality by classifying vitreous kernels, while reducing variability across 23 samples collected over four years. An industrial field test confirmed its applicability at scale. Genetic and environmental factors were also examined, revealing that varietal differences and flowering time were not significant contributors to variations in vitreousness. The findings suggest that color sorting is a reliable tool for enhancing wheat quality until more environmentally stable cultivars are developed, providing economic benefits through improved and consistent product quality.

Keywords: color sorting; kernel vitreousness; protein content; quality; wheat.

Figure 1

Differences in kernel vitreousness in relation to kernel hardness and protein content. (A) Mixed vitreous and starchy kernels observed in Korean hard white wheat cv. Baekkang. The solid circle represents vitreous kernels, while the dashed circle represents starchy kernels. (B, C) Hand-sorted vitreous kernels. (D, E) Hand-sorted starchy kernels. (F) Significant differences in kernel hardness between vitreous and starchy kernels (P < 0.0001). (G) Significant differences in flour protein contents between vitreous and starchy kernels (P < 0.0001).

Figure 2

Classification of vitreous kernels from starchy kernels using a color sorter. (A) An image of hard white wheat cv. Keumgang from an RGB converter program. (B) A diagram of classifying wheat kernels into four groups based on kernel vitreousness, using a color sorter.

Figure 3

Different kernel traits between groups classified with the color-sorting system. Difference in kernel lightness among (A) cv. Keumgang, (B) cv. Jokyoung, and (C) cv. Baekkang. Difference in kernel hardness among (D) cv. Keumgang, (E) cv. Jokyoung, and (F) cv. Baekkang. Difference in flour protein contents between groups in (G) cv. Keumgang, (H) cv. Jokyoung, and (I) cv. Baekkang. Different letters on plots indicate significant differences between each other (P <0.01).

Figure 4

Application of the color-sorting system on 23 wheat samples collected from 10 regions in Korea over four years. Different kernel colors observed between groups in (A) hard white wheat cv. Jokyoung and (B) hard red wheat cv. Hwanggeuaml. Variation in (C) protein content, (D) dry gluten content, and (E) SDS sedimentation values among groups. Different letters on the plots indicate significant differences (P < 0.01). (F) PCA analysis on protein content (PC), dry gluten content (DGC), gluten index (GI), and SDS-sedimentation value (SDSS) categorized by sorted groups. Increased kernel vitreousness in cv. Baekkang resulted in high (G) protein content, (H) SDS sedimentation value, (I) loaf volume, and (J) low bread firmness (*** P < 0.001).

Figure 5

Field demonstration test of the color sorting system in an industrial setting. (A) Layout of the grain cleaning and sorting process. (B) A simplified diagram of classifying wheat kernels into three groups based on kernel vitreousness, using the color sorter. (C) Proportion of each grade in hard white wheat cv. Baekkang and hard red wheat cv. Hwanggeumal. Differences in protein contents among groups in (D) cv. Baekkang and (E) cv. Hwanggeumal (P < 0.01).

Figure 6

Reason for differences in kernel vitreousness in the same cultivar in the same field according to flowering time. (A) Tagging stems according to the heading time. (B) Tagging in field conditions. (C) Dividing a spike into three sections. (D) Network plot of a correlation between grain characteristics and protein content. L: kernel lightness; DTH: days to heading; b: b value; PC: flour protein content; TGW: thousand-grain-weight. Differences in kernel lightness between tillers and spikelet order in (E) cv. Keumgang, (F) cv. Jokyoung, and (G) cv. Hwanggeumal. Differences in flour protein contents between tillers and spikelets order in (H) cv. Keumgang, (I) cv. Jokyoung, and (J) cv. Hwanggeumal. Different letters indicate significant difference (P < 0.01).