Development of a machine vision system for the determination of some of the physical properties of very irregular small biomaterials

nique is presented for the volume estimation of very irregular small biomaterials (wheat and rice-paddy grains). Two common cylindrical small biomaterials, the Alvand variety of wheat grain and the Neda variety of paddy grain were considered for examination. The captured images were exported to be processed by an image processing software (ImageJ) and the edge-extracted image was used in SolidWorks for the 3D reconstruction of the model. The revolved images in the SolidWork were used to estimate the volume of the examined grains. The estimated volume was then compared with the conventional mathematical expression and also with the real volume measurement using the fluid displacement method. Volume estimation using machine vision and image processing techniques has a considerably lower mean error (9.5%) in comparison to the mathematical error (14.7%). The average value of cylindricity for Alvand wheat was found to be equal to 82.34% at a moisture content of 11.83%. The new cylindricity factor had a significantly smaller standard deviation in comparison to the standard deviation of the sphericity factor for the examined cylindrical crops (61.5% for the wheat grains and 59.6% for the paddy grains). The new cylindricity factor can be used for the heat and mass transfer modelling of cylindrical crops. K e y w o r d s: machine vision, physical properties, volume, image processing, ImageJ, sphericity INTRODUCTION Huge postharvest losses occur in the handling and processing of high-quality food products and biomaterials due to the non-optimized design of certain equipment (Narendra et al., 2010). Cereals are one of the most prominent nutrition materials in Iranian food consumption. More than 50 percent of the nutrition material for each Iranian is provided by different kinds of breads (Mirasi et al., 2014). Therefore, a knowledge of the physical properties of the cereals is necessary for manufacturing optimized devices for handling, cleaning, conveying, storing and milling (Unal, 2009; Kalantari, 2016; Kalantari and Eshtavad, 2013). An understanding of the aeroand hydrodynamic properties of agricultural products is required for the movement of biomaterials by water or air and the separation of foreign materials from them. The density, shape and drag coefficient are the physical properties required to calculate the terminal velocity of an object in a fluid. In air transport or pneumatic separation, the air velocity is greater than the terminal velocity of the object. Moreover, for small biomaterials such as grains to descend slowly, the air velocity must be slightly lower than the terminal velocity. © 2022 Institute of Agrophysics, Polish Academy of Sciences D. KALANTARI et al. 28 Therefore, the shape of the biological material, i.e., spherical or cylindrical, has a very important role to play in grain flow through the discharge gates of silos and grain drills, as well as determining the air flow around the grains during movement which in turn allows for the determination of the drag coefficient (Chen et al., 2020; Kalantari, 2016; Dziki and Laskowski, 2005). For most cereals, the bulk porosity is between 35 and 55%; therefore the porous nature of the grain mass provides the possibility that in the blowing and drying process, almost all of the grains are in contact with air (Navarro and Noyes, 2002). Another important physical property of agricultural crops and materials is volume, which is an important parameter in the mass transfer and heat transfer of these biomaterials. Several mathematical expressions have been proposed for computing and estimating the volume of irregular materials (e.g., Kheiralipour et al., 2008; Unal 2009; El Fawal et al., 2009). As an example, the temperature rise of a biomaterial, e.g., a cereal grain in a microwave field is equal to Eq. (1) (Jafari et al., 2018):