Nondestructive diagnostics of magnesium deficiency based on distribution features of chlorophyll concentrations map on cucumber leaf

Shi, Jiyong; Li, Wenting; Zhai, Xiaodong; Guo, Zhiming; Holmes, Mel; Tahir, Haroon Elrasheid; Zhang, Xiaobo

doi:10.1080/01904167.2019.1659332

Cited by 10 publications

(4 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Remote sensing, specifically proximal sensing, can provide an effective alternative in assisting nutritional analysis of plants more accurately. The use of proximal sensors has an advantage over traditional agronomic methods since it allows to infer vegetation conditions in a non-invasive and non-destructive manner [32][33][34][35].…”

Section: Introductionmentioning

confidence: 99%

A Machine Learning Framework to Predict Nutrient Content in Valencia-Orange Leaf Hyperspectral Measurements

et al. 2020

View full text Add to dashboard Cite

This paper presents a framework based on machine learning algorithms to predict nutrient content in leaf hyperspectral measurements. This is the first approach to evaluate macro- and micronutrient content with both machine learning and reflectance/first-derivative data. For this, citrus-leaves collected at a Valencia-orange orchard were used. Their spectral data was measured with a Fieldspec ASD FieldSpec® HandHeld 2 spectroradiometer and the surface reflectance and first-derivative spectra from the spectral range of 380 to 1020 nm (640 spectral bands) was evaluated. A total of 320 spectral signatures were collected, and the leaf-nutrient content (N, P, K, Mg, S, Cu, Fe, Mn, and Zn) was associated with them. For this, 204,800 (320 × 640) combinations were used. The following machine learning algorithms were used in this framework: k-Nearest Neighbor (kNN), Lasso Regression, Ridge Regression, Support Vector Machine (SVM), Artificial Neural Network (ANN), Decision Tree (DT), and Random Forest (RF). The training methods were assessed based on Cross-Validation and Leave-One-Out. The Relief-F metric of the algorithms’ prediction was used to determine the most contributive wavelength or spectral region associated with each nutrient. This approach was able to return, with high predictions (R2), nutrients like N (0.912), Mg (0.832), Cu (0.861), Mn (0.898), and Zn (0.855), and, to a lesser extent, P (0.771), K (0.763), and S (0.727). These accuracies were obtained with different algorithms, but RF was the most suitable to model most of them. The results indicate that, for the Valencia-orange leaves, surface reflectance data is more suitable to predict macronutrients, while first-derivative spectra is better linked to micronutrients. A final contribution of this study is the identification of the wavelengths responsible for contributing to these predictions.

show abstract

Section: Introductionmentioning

confidence: 99%

A Machine Learning Framework to Predict Nutrient Content in Valencia-Orange Leaf Hyperspectral Measurements

et al. 2020

View full text Add to dashboard Cite

show abstract

“…In addition to containing beneficial information, raw spectral data contain useless information and noise interference information, such as the baseline drift and high-frequency noise, due to the effect of nonsample information, such as environmental information and machine operating conditions. Therefore, the preprocessing of raw spectral data is critical for removing useless information and improving the accuracy and stability of modeling ( Shi et al, 2019 ). In this study, the best spectral preprocessing method was selected from among standard normal variable transformation, the Savitzky–Golay (SG) method, vector normalization, the first-derivative method, the second-derivative method, and multiple scatter correction.…”

Section: Methodsmentioning

confidence: 99%

Application of spectral features for separating homochromatic foreign matter from mixed congee

et al. 2021

Self Cite

View full text Add to dashboard Cite

Highlights A method that can separate homochromatic FM in mixed congee was proposed. Spectral features of FM and mixed congee were extracted to build recognition model. The SVM model achieved high identification rates (99.17%) for homochromatic FM. The proposed method is better than computer vision in separating homochromatic FM.

show abstract

“…In addition, in this study, the nine characteristic wavelengths extracted by the UVE-CARS combined algorithms were consistent with previous studies reported (Riza, Suzuki, Ogawa, & Kondo, 2017;Workman, 1996), 967-1,095 nm was related to O H bond stretching vibrations, and 1,165-1,390 nm was related to C H bond stretching vibrations, 1,620-1,800 nm was related to the first overtone of C H stretching vibration. Therefore, the selected character- Previous studies have explored the feasibility of HSI of predicting chlorophyll concentration distribution map (including nitrogen and magnesium deficiency by cucumber leaf) in specific types of cucumber leaf within a single spectral range, such as Vis-NIR (Shi et al, 2019;Zou et al, 2010;Zou et al, 2011) and NIR region (Shi et al, 2012), and it is all predicted by linear models. However, in this work, the linear MLR and nonlinear LS-SVM were considered and compared.…”

Section: Modeling and Comparative Analysismentioning

confidence: 99%

Nondestructive prediction and visualization of total flavonoids content in Cerasus Humilis fruit during storage periods based on hyperspectral imaging technique

Bin

Zhang

2021

J Food Process Engineering

View full text Add to dashboard Cite

In this study, the changes of total flavonoids content in Cerasus Humilis fruits during storage periods were detected by using hyperspectral imaging (HSI) technique over the spectral region of 895-1,700 nm, and the spatial distribution of flavonoids was visualized during 24 days storage periods. The hyperspectral images of 240 samples of Cerasus Humilis were collected at different storage times (0, 8, 16, and 24 days). The Monte-Carlo outlier detection method was applied to identify four abnormal samples, three spectral preprocessing methods were used to preprocess the original spectral data, including savitzky-golay, standard normal variate, and baseline correction (BC). The x-loading weights, competitive adaptive reweighed sampling (CARS), uninformative variable elimination (UVE), UVE-CARS and the combination of UVE and successive projections algorithm (UVE-SPA) were used to reduce the dimension of spectral data and extract the characteristic wavelength, respectively. Then multiple linear regression and nonlinear least squared support vector machine (LS-SVM) regression models were developed based on full-spectrum data and selected characteristic wavelength.The results showed that BC was the best preprocessing approach, the LS-SVM models based on the nine effective wavelengths selected by UVE-CARS achieved the best results with correlation coefficient of prediction of .9357, root mean square error of prediction of 2.0107, and residual prediction deviation of 2.2809, respectively. The overall results demonstrated that the HSI technology coupled with chemometric algorithms is feasible to determine total flavonoids content, UVE-CARS-LS-SVM model was the optimal model. Visualization of the total flavonoids content during storage periods distribution map was performed, which clearly showed that the total flavonoids content kept decreasing with storage time. Practical ApplicationsCerasus Humilis is a fruit tree resource endemic to China. The content of total flavonoids is one of the key factors that affect the edible quality of Cerasus Humilis fruit.With the increase of storage time, the active components of flavonoids are easy to be degraded by oxidation, the quality and commercial value of Cerasus Humilis fruit were negatively affected. Thus, the detection of total flavonoids content during the

show abstract

Nondestructive diagnostics of magnesium deficiency based on distribution features of chlorophyll concentrations map on cucumber leaf

Cited by 10 publications

References 29 publications

A Machine Learning Framework to Predict Nutrient Content in Valencia-Orange Leaf Hyperspectral Measurements

A Machine Learning Framework to Predict Nutrient Content in Valencia-Orange Leaf Hyperspectral Measurements

Application of spectral features for separating homochromatic foreign matter from mixed congee

Nondestructive prediction and visualization of total flavonoids content in Cerasus Humilis fruit during storage periods based on hyperspectral imaging technique

Contact Info

Product

Resources

About