canada produces around 57 Mt of grains every year and about 46yo ofthat is exported. wheat is the major grain in canada, and the annual production is about 43% (24.5 Mt) of the total grain production. To minimize the losses by stored-product insects, early detection of infestations is required to camy out control measures. Knowledge on heating behavior and quality changes during drying using different methods such as microwave heating are essential to develop alternative grain drying systems. similarly, in grain handling facilities, quick, reliable machine vision methods are needed to assist the grain inspectors in the determination of grain grading factors such as class identification. Therefore, the potential uses of an infrared thermal imaging method were studied for five similar applications in grain handling operations.

An un-cooled focal planar array type infrared thermal camera (Model:​ ThermaCAM™ SC500 of FLIR systems, Burlington, ON, Canada;​ spectral range:​ 7.5 to 13.0 μm), with 320×240 pixels, was used to take thermal images of the grain bulk samples' Fol single kernel analysis, a 50 μm close-up lens was attached to the original lens of the camera to get magnified thermal images.

Non-uniform heating patterns were observed on the surface of barley, canola and wheat after microwave treatment in a pilbt-scale drier. The difference between maximum and minimum temperatures (aT) on the surface of the grain bulk were 62.g to 69.5°C, 64.3 to 75.6°C and 39.5 to 59.2oc for wheat, barley and canola, respectively, when the applied microwave power and exposure time were 500 W and 56 s. Germination percentages and fat acidity value (FAV) were determined for the wheat samples collected from the high temperature and normal temperature regions of bulk grain after microwave treatment. The germination percentages were signifìcantly (α=0.05) lower for samples collected from high temperature region than those from the normal temperature zones. Similarly, the FAV was significantly different for the samples collected frorn the high temperature and normal temperature regions.

The capability of a thermal imaging method to detect a hot spot in a stored grain silo was determined. The hot spot was detected from the thermal images of the silo_wall and grain bulk (as a high temperature region) when it was located at 0.3 m from the silo wall and below the grain surface, respectively. Hot spots were not detected in thermal images during wind and cold weather conditions. Thermal imaging can not be used as an independent method to detect the hot spot in a silo by monitoring the surface temperature of the silo wall.

A thermal imaging system was developed to identify eight western Canadian wheat classes using bulk sample (20 g) analyses. The overall classification accuracies of a quadratic discriminant method with 8 classes mixed, red classes mixed (4 classes), white classes mixed (4 classes) and pairwise comparison (2 classes) were;​ 76,87,79 and 95%, and, 64, 87, 77 and, 91% using bootstrap and leave-one-out validation techniques, respectively.

The efficiency of a thermal imaging system was determined to detect the presence of Cryptolestes ferrugineus (Stephens) inside wheat kernels at six developmental stages (four larval, pupal and adult). The overall classification accuracy for a quadratic discriminant method was 83.5 and17.7Yo for infested and sound kernels, respectively, and for a linear discriminant method it was 17.6 and 83.0% for infested and sound kernels, respectively.