Identification of distinguishable engineering characteristics of black mustard (Brassica nigra), brown mustard (Brassica juncea) and argemone (Argemone mexicana)

Adulteration of edible oils is prevalent and widespread. Argemone (Argemone mexicana) seeds are a common adulterant to black mustard (Brassica nigra) or brown mustard (Brassica juncea) seeds. This paper reports a comparative analysis of the different engineering properties (viz. geometric, gravimetric, frictional, mechanical, chromatic and aerodynamic) of these seeds. Fisher's least significant difference test was applied to analyze the statistical difference between the mean values of the engineering properties of these seeds. Significant (p<0.01) difference was observed in weight, axial length, bulk density and terminal velocity of these seeds. The mechanical properties and frictional properties of these three seeds on plywood and aluminium surface varied significantly (p<0.05). Machine learning techniques, namely- artificial neural network, radial basic function, support vector machine and random forest were applied to classify B. juncea, B. nigra and A. mexicana seeds. Random forest classification model demonstrated highest accuracy (86.78%) with a kappa statistics of 0.80.