Technical Note: Wheat Class Identification Using Thermal Imaging: A Potential Innovative Technique.

An infrared thermal imaging system was assessed for its capability to identify eight western Canadian wheat classes. Twenty-gram wheat samples at 14% moisture content (wet basis) were spread in a 100 × 100 mm monolayer and heated by a plate heater (maintained at 90°C) placed at a distance of 100 mm from the grain layer. The surface temperatures of the top surface of the grain bulk were imaged before heating, after heating for 180 s, and after cooling for 30 s using a thermal camera (n = 100 samples in each class). Temperature rise (after heating) and drop (after cooling) were significantly different for wheat classes (a = 0.05). Wheat samples were classified with twelve derived thermal features using statistical classifiers. Overall classification accuracies of an eight-class model, red class model (four classes), white class model (four classes), and pairwise comparison (two-class model) using a quadratic discriminant method were 76%, 87%, 79%, and 95%, respectively. The thermal imaging approach may have potential to develop class identification methods in wheat handling facilities. However, further investigations are required to determine the performance of this system for wheat samples mixed with defects, with different moisture contents, and grown in different locations and years.