A Non-Destructive Moisture Detection System for Unshelled Green Tea Seed Kernels Based on Microwave Technology with Multi-Frequency Scanning Signals

A self-developed microwave moisture detection system (ranged from 2.00 GHz to 10.00 GHz) based on multi-frequency sweep technology was used to quickly determine the moisture content of tea seed kernels without breaking the shells. A multi-frequency evaluation method combined cross-validation and maj...

Full description

Saved in:
Bibliographic Details
Main Authors: Bo Zhou, Ye Yuan, Zhenbo Wei, Siying Li
Format: Article
Language:English
Published: MDPI AG 2025-02-01
Series:Sensors
Subjects:
Online Access:https://www.mdpi.com/1424-8220/25/5/1324
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:A self-developed microwave moisture detection system (ranged from 2.00 GHz to 10.00 GHz) based on multi-frequency sweep technology was used to quickly determine the moisture content of tea seed kernels without breaking the shells. A multi-frequency evaluation method combined cross-validation and majority voting rules was proposed to select the optimal microwave features from the original microwave signals. Firstly, the moisture content of tea seed kernels was detected by the moisture detection system, and the determination coefficients of the ANN model established based on seven attenuations and six phase-shifts were over 0.999. Then, the moisture content of unshelled tea seeds was detected, and the determination coefficients of the ANN model based on 13 preferred frequency features were over 0.995. Moreover, the predicted moisture values of unshelled tea seeds were calibrated accurately by a moisture function (y = −0.017x<sup>2</sup> + 1.431x − 1.019). Above all, the self-developed system could achieve non-destructive moisture content prediction of tea seed kernels.
ISSN:1424-8220