Visible and near-infrared spectroscopy and deep learning application for the qualitative and quantitative investigation of nitrogen status in cotton leaves

Abstract

Leaf nitrogen concentration (LNC) is a critical indicator of crop nutrient status. In this study, the feasibility of using visible and near-infrared spectroscopy combined with deep learning to estimate LNC in cotton leaves was explored. The samples were collected from cotton’s whole growth cycle, and the spectra were from different measurement environments. The random frog (RF), weighted partial least squares regression (WPLS), and saliency map were used for characteristic wavelength selection. Qualitative models (partial least squares discriminant analysis (PLS-DA), support vector machine for classification (SVC), convolutional neural network classification (CNNC) and quantitative models (partial least squares regression (PLSR), support vector machine for regression (SVR), convolutional neural network regression (CNNR)) were established based on the full spectra and characteristic wavelengths. Satisfactory results were obtained by models based on CNN. The classification accuracy of leaves in three different LNC ranges was up to 83.34%, and the root mean square error of prediction (RMSEP) of quantitative prediction models of cotton leaves was as low as 3.36. In addition, the identification of cotton leaves based on the predicted LNC also achieved good results. These results indicated that the nitrogen content of cotton leaves could be effectively detected by deep learning and visible and near-infrared spectroscopy, which has great potential for real-world application.

Rapid and cost‐effective assessment of nutrients in pistachio (Pistacia vera L.) leaves through Fourier transform near‐infrared spectroscopy (FT‐NIRS)

“…It was the first study applied to pistachio tree leaves and also included the estimation of multiple macro-and microelements in contrast to studies with only one or several nutrient elements. Xiao et al (2022) tested an NIRS system and deep learning technique to assess leaf nitrogen concentration (LNC) of fresh cotton leaves, and they found a good accuracy of up to 83% to classify the leaves into three LNC groups. Phanomsophon et al (2022) showed that the NIRS method provided good accuracy (>80%) for the classification of N, P, and K concentration levels of durian fruit tree leaves and stated that they obtained higher accuracy with fresh leaf samples than dried-ground leaf samples.…”

“…Xiao et al. (2022) tested an NIRS system and deep learning technique to assess leaf nitrogen concentration (LNC) of fresh cotton leaves, and they found a good accuracy of up to 83% to classify the leaves into three LNC groups. Phanomsophon et al.…”

Nitrogen sensing and regulatory networks: it's about time and space

“…This is why an important trend is now rising as it relates to the measure of N content on plants from image analysis. Most of the time these approaches use multispectral images to classify the N content of different crops, including maize ( Nguyen et al 2023 ; Wijewardane et al 2023 ), cotton ( Xiao et al 2022 ), or sorghum ( Wijewardane et al 2023 ).…”