Identification of Soybean Seed Varieties Based on Hyperspectral Imaging Technology.

Shaolong Zhu,Jinlong Zhang,Maoni Chao,Xinjuan Xu,Puwen Song,Zhongwen Huang,Jinyu Zhang

doi:10.3390/s19235225

Shaolong Zhu, Jinlong Zhang + Show 5 more

Open Access

https://doi.org/10.3390/s19235225

Copy DOI

Abstract

Hyperspectral imaging is a nondestructive testing technology that integrates spectroscopy and iconology technologies, which enables us to quickly obtain both internal and external information of objects and identify crop seed varieties. First, the hyperspectral images of ten soybean seed varieties were collected and the reflectance was obtained. Savitzky-Golay smoothing (SG), first derivative (FD), standard normal variate (SNV), fast Fourier transform (FFT), Hilbert transform (HT), and multiplicative scatter correction (MSC) spectral reflectance pretreatment methods were used. Then, the feature wavelengths and feature information of the pretreated spectral reflectance data were extracted using competitive adaptive reweighted sampling (CARS), the successive projections algorithm (SPA), and principal component analysis (PCA). Finally, 5 classifiers, Bayes, support vector machine (SVM), k-nearest neighbor (KNN), ensemble learning (EL), and artificial neural network (ANN), were used to identify seed varieties. The results showed that MSC-CARS-EL had the highest accuracy among the 90 combinations, with training set, test set, and 5-fold cross-validation accuracies of 100%, 100%, and 99.8%, respectively. Moreover, the contribution of spectral pretreatment to discrimination accuracy was higher than those of feature extraction and classifier selection. Pretreatment methods determined the range of the identification accuracy, feature-selective methods and classifiers only changed within this range. The experimental results provide a good reference for the identification of other crop seed varieties.

Highlights

Seed varieties are directly related to the yield and quality of soybeans
Pretreatment methods determined the range of the identification accuracy, feature-selective methods and classifiers only changed within this range
The correlation coefficients were determined by comparing the spectral reflectivity of each band with the crude protein content and crude fat content (Figure 3c), and the highest value was 0.34 at 749 nm, indicating that the spectral reflectance of different varieties is not significantly related to the crude protein content or crude fat content

Summary

Introduction

Seed varieties are directly related to the yield and quality of soybeans. As people’s requirements for food quality become increasingly higher, it is necessary to process different uses according to different seed varieties. The soymilk and tofu made by high-protein soybeans are more delicious [2,3]. The identification of seed varieties is an urgent problem to be solved in agricultural production, seed sales, and food processing. The common methods of seed identification in China and abroad include morphological methods, the gel electrophoresis of soluble seed proteins [4,5], direct analysis with real-time mass spectrometry [6], isoenzyme electrophoresis [7], liquid chromatography [8], and simple sequence repeat (SSR) analysis [9]

Objectives

Methods

Results

Conclusion