Assessing the intrapore volume and surface area of single corn kernel using CT scan imaging and machine vision

The intragranular porosity of granular materials is crucial across various fields and influences mechanical strength, permeability, moisture retention, and overall reactivity. However, the existing pore structure assessment techniques are complicated, labor-intensive, and time-consuming, while the use of digital imaging and analysis is promising. Therefore, a machine vision ImageJ plugin was developed and tested that used digital computed tomography (CT) scans of single corn for the assessment of intrapore (within grain) volume and surface area as a proof-of-concept. Among the exhibited two distinct intrapore groups, the top has a higher number of intrapores ( ), while the sum of areas and perimeters in the bottom is greater. Compared to the whole kernel, the sum of areas of intrapores is minimal ( 27 X smaller), yet their sum of perimeters is significantly larger ( 2.77 X greater). The intrapore sizes are much smaller ( – ) compared to the whole kernel ( ). Shape descriptors indicate that the intrapores are generally elongated (aspect ratio = ; circularity = ). The total volume of intrapores ( ) is only of the total volume of the whole kernel ( ); however, the total intrapores' surface area ( ) is of the whole kernel's surface area ( ) due to the phenomenon of new surface area generation. The trapezoidal method of integration over Simpson's is recommended because of its simplicity for volume, and the conical frustum method for surface area determination. The developed plugin produced rapid analysis (CPU time for image slices), and the methodology can be easily extended to other grains and agricultural products to inspect the overall internal quality, damage, or structure. • Machine vision ImageJ plugin for corn intrapore volume and surface area was developed. • Total volume of intrapores is much less, while surface area is higher than whole kernel. • Shape descriptors indicate that the intrapores are generally slightly elongated. • Trapezoidal method of integration is recommended for volume due to its simplicity. • Plugin produced rapid calculations and validated with available reported results.