Detection for lead pollution level of lettuce leaves based on deep belief network combined with hyperspectral image technology

Sun, Jun; Cao, Yan; Zhou, Xin; Wu, Minmin; Sun, Yidan; Hu, Yinghui

doi:10.1111/jfs.12866

Cited by 18 publications

(9 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It can be seen that when the the concentration of Pb stress in rape leaves was 100 mg/L, the SPAD value was higher than that of the ck group without Pb stress, indicating that the low concentration of Pb can increase the SPAD value and promote the production of chlorophyll. When the concentration of Pb stress was more than 100 mg/L, SPAD value decreased with the increase of Pb content, indicating that excessive Pb concentration in leaves can inhibit the production of chlorophyll (Sun et al, 2021). With the increase of chlorophyll content in rape leaves, photosynthesis was enhanced, and the absorption of red and blue light was increased, leading to the gradual decrease of its reflectivity (Fang, Song, Cao, He, & Qiu, 2007).…”

Section: Resultsmentioning

confidence: 99%

“…The absorption of Pb by plants can cause changes in chlorophyll content of leaves, affecting the spectral reflectance of visible light band. In addition, the structure of leaf membrane system can be destroyed by Pb stress, and the reflectance of near‐infrared band can be changed (Liang, Shi, Ma, Xing, & Yu, 2010; Sun et al, 2021). Therefore, the Pb content of rape leaves can be reflected by analyzing the spectral data of hyperspectral.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Nondestructive detection of lead content in oilseed rape leaves based on MRF‐HHO‐SVR and hyperspectral technology

Cao

Sun

Yao

et al. 2021

J Food Process Engineering

Self Cite

View full text Add to dashboard Cite

Lead (Pb) pollution poses a huge threat to environmental quality and food safety. Eating products made by excessive Pb contaminated oilseed rape can seriously harm human health. To detect Pb content in rape leaves more accurately, a method based on visible‐near infrared (400.68–1000.61 nm) hyperspectral technology was studied. First, the first derivative was used to pretreat the original spectral data. To reduce the number of iterations and the randomness of variable selection, the random frog (RF) algorithm was improved and named modified random frog (MRF). Then, the characteristic wavelengths were selected by MRF and competitive adaptive reweighted sampling (CARS), respectively, and support vector regression (SVR) model was established to predict the Pb content in rape. MRF was determined as the best feature selection algorithm. Finally, Harris Hawks Optimizer (HHO) was used to optimize SVR and the coefficient of determination (R2) and root mean square error (RMSE) of the prediction set were 0.9431 and 0.1645 mg/kg, respectively. Therefore, the combination of hyperspectral technology and the optimal model (MRF‐HHO‐SVR) is feasible for nondestructive detection of Pb content in rape leaves. Practical applications In recent years, the development of mineral resources has caused more and more serious pollution of heavy metals in the soil. The problem of lead pollution is particularly prominent, and various crops have been polluted to varying degrees. The physiological indicators and nutrient content of Pb contaminated oilseed rape have obvious changes. Excessive consumption of lead contaminated rape products can be hazardous to human health. A method for detecting lead content in rape leaves based on hyperspectral technology was studied in this article. The results show that the hyperspectral technology could be used for the accurate and nondestructive determination of lead content in rape leaves, which provides a new method and idea for the detection of lead content in crops.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Nondestructive detection of lead content in oilseed rape leaves based on MRF‐HHO‐SVR and hyperspectral technology

Cao

Sun

Yao

et al. 2021

J Food Process Engineering

Self Cite

View full text Add to dashboard Cite

show abstract

“…We selected 10 symmetrically distributed ROIs for each hyperspectral image. The average of the spectra in each ROI were used as sample spectra for that moment, and each moment contained 30 spectra, for a total of 180 spectra [ 16 ].…”

Section: Methodsmentioning

confidence: 99%

Nondestructive Testing and Visualization of Catechin Content in Black Tea Fermentation Using Hyperspectral Imaging

Dong

Yang

Liu

et al. 2021

Sensors

View full text Add to dashboard Cite

Catechin is a major reactive substance involved in black tea fermentation. It has a determinant effect on the final quality and taste of made teas. In this study, we applied hyperspectral technology with the chemometrics method and used different pretreatment and variable filtering algorithms to reduce noise interference. After reduction of the spectral data dimensions by principal component analysis (PCA), an optimal prediction model for catechin content was constructed, followed by visual analysis of catechin content when fermenting leaves for different periods of time. The results showed that zero mean normalization (Z-score), multiplicative scatter correction (MSC), and standard normal variate (SNV) can effectively improve model accuracy; while the shuffled frog leaping algorithm (SFLA), the variable combination population analysis genetic algorithm (VCPA-GA), and variable combination population analysis iteratively retaining informative variables (VCPA-IRIV) can significantly reduce spectral data and enhance the calculation speed of the model. We found that nonlinear models performed better than linear ones. The prediction accuracy for the total amount of catechins and for epicatechin gallate (ECG) of the extreme learning machine (ELM), based on optimal variables, reached 0.989 and 0.994, respectively, and the prediction accuracy for EGC, C, EC, and EGCG of the content support vector regression (SVR) models reached 0.972, 0.993, 0.990, and 0.994, respectively. The optimal model offers accurate prediction, and visual analysis can determine the distribution of the catechin content when fermenting leaves for different fermentation periods. The findings provide significant reference material for intelligent digital assessment of black tea during processing.

show abstract

“…Recent advances in machine learning have shown that deep learning algorithms can automatically learn to extract features from raw data and significantly improve modeling performance for many spectral analysis tasks ( Singh et al, 2018 ; Kanjo et al, 2019 ; Rehman et al, 2020 ). Sun et al (2019 , 2021) employed a deep brief network to estimate cadmium and lead contents of lettuces with high accuracy. Rehman et al (2020) also developed a modified Inception module to predict the relative water content (RWC) of maize leaves and achieved a determination coefficient ( R 2 ) of 0.872 for RWC.…”

Section: Introductionmentioning

confidence: 99%

Hyperspectral Technique Combined With Deep Learning Algorithm for Prediction of Phenotyping Traits in Lettuce

Fan

et al. 2022

Front. Plant Sci.

View full text Add to dashboard Cite

The currently available methods for evaluating most biochemical traits of plant phenotyping are destructive and have extremely low throughput. However, hyperspectral techniques can non-destructively obtain the spectral reflectance characteristics of plants, which can provide abundant biophysical and biochemical information. Therefore, plant spectra combined with machine learning algorithms can be used to predict plant phenotyping traits. However, the raw spectral reflectance characteristics contain noise and redundant information, thus can easily affect the robustness of the models developed via multivariate analysis methods. In this study, two end-to-end deep learning models were developed based on 2D convolutional neural networks (2DCNN) and fully connected neural networks (FCNN; Deep2D and DeepFC, respectively) to rapidly and non-destructively predict the phenotyping traits of lettuces from spectral reflectance. Three linear and two nonlinear multivariate analysis methods were used to develop models to weigh the performance of the deep learning models. The models based on multivariate analysis methods require a series of manual feature extractions, such as pretreatment and wavelength selection, while the proposed models can automatically extract the features in relation to phenotyping traits. A visible near-infrared hyperspectral camera was used to image lettuce plants growing in the field, and the spectra extracted from the images were used to train the network. The proposed models achieved good performance with a determination coefficient of prediction (Rp2) of 0.9030 and 0.8490 using Deep2D for soluble solids content and DeepFC for pH, respectively. The performance of the deep learning models was compared with five multivariate analysis method. The quantitative analysis showed that the deep learning models had higher Rp2 than all the multivariate analysis methods, indicating better performance. Also, wavelength selection and different pretreatment methods had different effects on different multivariate analysis methods, and the selection of appropriate multivariate analysis methods and pretreatment methods increased more time and computational cost. Unlike multivariate analysis methods, the proposed deep learning models did not require any pretreatment or dimensionality reduction and thus are more suitable for application in high-throughput plant phenotyping platforms. These results indicate that the deep learning models can better predict phenotyping traits of plants using spectral reflectance.

show abstract

Detection for lead pollution level of lettuce leaves based on deep belief network combined with hyperspectral image technology

Cited by 18 publications

References 24 publications

Nondestructive detection of lead content in oilseed rape leaves based on MRF‐HHO‐SVR and hyperspectral technology

Nondestructive detection of lead content in oilseed rape leaves based on MRF‐HHO‐SVR and hyperspectral technology

Nondestructive Testing and Visualization of Catechin Content in Black Tea Fermentation Using Hyperspectral Imaging

Hyperspectral Technique Combined With Deep Learning Algorithm for Prediction of Phenotyping Traits in Lettuce

Contact Info

Product

Resources

About