Submitted to: Journal of Agricultural Machinery
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: September 10, 2009
Publication Date: December 23, 2009
Citation: Yang, I., Delwiche, S.R., Kim, M.S., Tsai, C., Lo, Y. 2009. Determination of wheat kernel black point damage using hyper-spectral imaging. Journal of Agricultural Machinery. 18:29-44. Interpretive Summary: In the United States, wheat that undergoes official inspection, such as that for export, is assessed for its level of damaged kernels. The greater the damage, the lower will be the grade, quality, and trade value of the lot. Black point or black tip is a physiological condition that causes discoloration in kernels and is a contributor to kernel damage. As an alternative to human visual inspection (the traditional standard procedure), we explored the possibility of using fluorescence imaging to accentuate the contrast between black point and sound kernel surfaces. Using just one fluorescence wavelength in the region of green light, we were successful at identifying black point damaged kernels at better than 90 percent accuracy. These findings, when coupled with similar image analysis procedures for other forms of damage (heat-, frost-, mold-, and insect-), will be of potential benefit to official inspection offices and commercial processors.
Technical Abstract: A feasibility study was conducted on the use of hyperspectral imaging to differentiate sound wheat kernels from those with a damage condition called black point or black tip. Individual kernels of hard red spring wheat were loaded in indented slots on a blackened machined aluminum plate. Damage conditions, determined by official (USDA) inspection, were either sound (no damage) or damaged by the black tip condition alone. Hyperspectral imaging was separately performed under modes of reflectance from white light illumination and fluorescence from UV light (365 nm) excitation. Through analysis of wavelength images, one fluorescence wavelength (531 nm) was selected for image processing and classification analysis. Results indicated that with this wavelength alone, classification accuracy can be as high as 95% when kernels are oriented with their dorsal side toward the camera. It is suggested that improvement in classification can be made through the inclusion of multiple wavelength images.