The use of polarized light and image analysis in evaluations of the severity of fungal infection in barley grain

摘要

Barley grain infected with fungi of the genus Fusarium was analyzed with the use of a machine vision system. Mold-affected, black-tipped and healthy kernels were viewed under polarized light and non-polarized light as the reference method, and the resulting images were compared. Texture attributes were calculated in selected regions of interest (ROIs) and used in statistical modeling. The results were analyzed with the use of various pattern recognition methods and artificial neural networks. Mold-affected and black-tipped grains were more effectively separated from healthy kernels under polarized light, and this identification technique was 20-30% to 50% more accurate than the reference method.