Dielectric-based calibration algorithm for rapid and nondestructive determination of multiple quality attributes of in-shell nuts

Authors: Trabelsi, Samir; Lewis, Micah

Submitted to: IEEE Sensors Letters
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 1/23/2025
Publication Date: N/A
Citation: N/A

Interpretive Summary: In-shell nuts industry is a multi-billion dollar industry that is a growing rapidly with increasing demand worldwide. Currently, determining their quality attributes, dollar value, and conditions for their safe storage rely on tedious and time-consuming grading processes which utilize multiple outdated equipment that require the involvement of several well-trained inspectors. For instance, peanut grading protocol is a 50-year old protocol that involves determining the foreign material content, size, splits and damaged kernels, peanut “meat” content, kernel moisture content, and aflatoxin level. Several inspectors are involved in this process, and they make use of outdated mechanical means, visual inspection, and microscopy to complete the grading process. In this paper, a novel microwave sensing approach is proposed for determining instantaneously and nondestructively several quality attributes with the same low-cost microwave. In this novel approach, a dielectric-based algorithm was developed and allows the determination of three important quality attributes of in-shell nuts, namely, bulk density, moisture content (pods and kernels) and "meat" content from measurement of their dielectric properties at a single microwave frequency. For purpose of illustration, measurements were performed on in-shell peanuts with a low-cost microwave sensor, develop within our research group, and operating at 5.8 GHz. Results show that bulk density, moisture content (pods and kernels), and meat content can be predicted with standard errors of calibration that remain within the desired range of accuracy.

Technical Abstract: A dielectric-based calibration algorithm is proposed for simultaneous and nondestructive determination of three quality attributes of in-shell nuts, namely, bulk density, moisture content (pods and kernels), and “meat” content from measurement of their dielectric properties at a single microwave frequency. The calibration algorithm relies on the existence of direct correlations between the quality attributes and the dielectric properties measured at a single microwave frequency. Performance of this algorithm is demonstrated through measurements on in-shell peanuts with a low-cost microwave sensor operating at 5.8 GHz. Calibration equations correlating each quality attribute with in-shell peanuts dielectric properties are given along with the standard error of calibration (SEC).