Identification of maize seed varieties based on stacked sparse autoencoder and near‐infrared hyperspectral imaging technology

Abstract

Maize seed variety identification is the key to improving the quality and yield of maize. The study aimed to investigate a stacked sparse autoencoder combined with a cuckoo search (CS) optimized support vector machine (SSAE-CS-SVM) to meet the identification requirements of accurate detection. First, the near-infrared (NIR) (871.61-1766.32 nm) hyperspectral data of maize seeds were processed using Savitzky-Golay (SG) combined with standard normal variables (SNV). Subsequently, SSAE, SAE, principal component analysis (PCA), and competitive adaptive reweighted sampling were employed for feature extraction. Finally, the recognition model was constructed using the SoftMax regression model, SVM, and CS optimized SVM (CS-SVM). The results indicated that the SSAE-CS-SVM model achieved satisfactory performance (the training set and testing set accuracies were 99.45% and 95.81%, respectively). This study confirmed the great potential of combining NIR hyperspectral imaging technology with deep learning algorithms for maize seed variety identification. Practical applicationsThe identification of maize seed varieties is extremely important for ensuring seed purity and improving maize quality and yield. A method based on SSAE and NIR hyperspectral imaging technology was studied to identify maize seed varieties. The method proposed in this paper could identify maize seed varieties non-destructively and accurately, which provided a new way for the online detection of seed varieties.

Exploring the Steps of Infrared (IR) Spectral Analysis: Pre-Processing, (Classical) Data Modelling, and Deep Learning

“…They made this to study near-infrared data from maize seeds. Their goal was to correctly identify different types of maize seeds [104]. Due to the limited training samples for NIR data, GAN is used to create more training samples [105].…”

A Review of Machine Learning for Near-Infrared Spectroscopy

“…Wojciech et al at the AGH University of Science and Technology employed three different models to evaluate soil and land: an SAE, a CNN, and the stack model, which consists of a collection of multilayer perceptron algorithms with two distinct methods for regression estimation that analyze the Vis-NIR spectral response signal [123]. Fu et al proposed a stacked sparse autoencoder combined with a cuckoo search (CS)-optimized-support vector machine (SSAE-CS-SVM) for analyzing the near-infrared (NIR) hyperspectral data of maize seeds (871.61-1766.32 nm) to achieve maize seed variety identification [124]. Jun.…”

Determination of the ADF and IVOMD Content of Sugarcane Using Near Infrared Spectroscopy Coupled with Chemometrics

“…Although it has been used in many different areas, the most intensive studies have been carried out in the food sector. Several studies have been conducted to determine food quality (Mohamed et al 2021;Teye et al 2020;Yang et al 2022), detect food adulteration (De Girolamo et al 2020;Genis et al 2021;Laborde et al 2021), and classify food types (Fu et al 2022).…”