Support Vector Machine and Artificial Neural Network Models for the Classification of Grapevine Varieties Using a Portable NIR Spectrophotometer

Analytical Chemistry Strategies in the Use of Miniaturised NIR Instruments: An Overview

Miniaturized NIR instruments have been increasingly used in the last years, and they have become useful tools for many applications on a broad variety of samples. This review focuses on miniaturized NIR instruments from an analytical point of view, to give an overview of the analytical strategies used in order to help the reader to set up their own analytical methods, from the sampling to the data analysis. It highlights the uses of these instruments, providing a critical discussion including current and future trends.

Design and Experiment of a Portable Near-Infrared Spectroscopy Device for Convenient Prediction of Leaf Chlorophyll Content

This study designs a spectrum data collection device and system based on the Internet of Things technology, aiming to solve the tedious process of chlorophyll collection and provide a more convenient and accurate method for predicting chlorophyll content. The device has the advantages of integrated design, portability, ease of operation, low power consumption, low cost, and low maintenance requirements, making it suitable for outdoor spectrum data collection and analysis in fields such as agriculture, environment, and geology. The core processor of the device uses the ESP8266-12F microcontroller to collect spectrum data by communicating with the spectrum sensor. The spectrum sensor used is the AS7341 model, but its limited number of spectral acquisition channels and low resolution may limit the exploration and analysis of spectral data. To verify the performance of the device and system, this experiment collected spectral data of Hami melon leaf samples and combined it with a chlorophyll meter for related measurements and analysis. In the experiment, twelve regression algorithms were tested, including linear regression, decision tree, and support vector regression. The results showed that in the original spectral data, the ETR method had the best prediction effect at a wavelength of 515 nm. In the training set, RMSEc was 0.3429, and Rc2 was 0.9905. In the prediction set, RMSEp was 1.5670, and Rp2 was 0.8035. In addition, eight preprocessing methods were used to denoise the original data, but the improvement in prediction accuracy was not significant. To further improve the accuracy of data analysis, principal component analysis and isolation forest algorithm were used to detect and remove outliers in the spectral data. After removing the outliers, the RFR model performed best in predicting all wavelength combinations of denoised spectral data using PBOR. In the training set, RMSEc was 0.8721, and Rc2 was 0.9429. In the prediction set, RMSEp was 1.1810, and Rp2 was 0.8683.

Fruit classification using attention-based MobileNetV2 for industrial applications

Recent deep learning methods for fruits classification resulted in promising performance. However, these methods are with heavy-weight architectures in nature, and hence require a higher storage and expensive training operations due to feeding a large number of training parameters. There is a necessity to explore lightweight deep learning models without compromising the classification accuracy. In this paper, we propose a lightweight deep learning model using the pre-trained MobileNetV2 model and attention module. First, the convolution features are extracted to capture the high-level object-based information. Second, an attention module is used to capture the interesting semantic information. The convolution and attention modules are then combined together to fuse both the high-level object-based information and the interesting semantic information, which is followed by the fully connected layers and the softmax layer. Evaluation of our proposed method, which leverages transfer learning approach, on three public fruit-related benchmark datasets shows that our proposed method outperforms the four latest deep learning methods with a smaller number of trainable parameters and a superior classification accuracy. Our model has a great potential to be adopted by industries closely related to the fruit growing and retailing or processing chain for automatic fruit identification and classifications in the future.

Rapid quantification of dissolved solids and bioactives in dried root vegetable extracts using near infrared spectroscopy

Artificial neural networks (ANN) were developed for prediction of total dissolved solids, polyphenol content and antioxidant capacity of root vegetables (celery, fennel, carrot, yellow carrot, purple carrot and parsley) extracts prepared from the (i) fresh vegetables, (ii) vegetables dried conventionally at 50 °C and 70 °C, and (iii) the lyophilised vegetables. Two types of solvents were used: organic solvents (acetone mixtures and methanol mixtures) and water. Near-infrared (NIR) spectra were recorded for all samples. Principal Component Analysis (PCA) of the pre-treated spectra using Savitzky-Golay smoothing showed specific grouping of samples in two clusters (1st: extracts prepared using methanol mixtures and water as the solvents; 2nd: extracts prepared using acetone mixtures as the solvents) for all four types of extracts. Furthermore, obtained results showed that the developed ANN models can reliably be used for prediction of total dissolved solids, polyphenol content and antioxidant capacity of dried root vegetable extracts in relation to the recorded NIR spectra.

Application of Visible/Infrared Spectroscopy and Hyperspectral Imaging With Machine Learning Techniques for Identifying Food Varieties and Geographical Origins

Food quality and safety are strongly related to human health. Food quality varies with variety and geographical origin, and food fraud is becoming a threat to domestic and global markets. Visible/infrared spectroscopy and hyperspectral imaging techniques, as rapid and non-destructive analytical methods, have been widely utilized to trace food varieties and geographical origins. In this review, we outline recent research progress on identifying food varieties and geographical origins using visible/infrared spectroscopy and hyperspectral imaging with the help of machine learning techniques. The applications of visible, near-infrared, and mid-infrared spectroscopy as well as hyperspectral imaging techniques on crop food, beverage, fruits, nuts, meat, oil, and some other kinds of food are reviewed. Furthermore, existing challenges and prospects are discussed. In general, the existing machine learning techniques contribute to satisfactory classification results. Follow-up researches of food varieties and geographical origins traceability and development of real-time detection equipment are still in demand.

Review of the Effects of Grapevine Smoke Exposure and Technologies to Assess Smoke Contamination and Taint in Grapes and Wine

Grapevine smoke exposure and the subsequent development of smoke taint in wine has resulted in significant financial losses for grape growers and winemakers throughout the world. Smoke taint is characterized by objectional smoky aromas such as “ashy”, “burning rubber”, and “smoked meats”, resulting in wine that is unpalatable and hence unprofitable. Unfortunately, current climate change models predict a broadening of the window in which bushfires may occur and a rise in bushfire occurrences and severity in major wine growing regions such as Australia, Mediterranean Europe, North and South America, and South Africa. As such, grapevine smoke exposure and smoke taint in wine are increasing problems for growers and winemakers worldwide. Current recommendations for growers concerned that their grapevines have been exposed to smoke are to conduct pre-harvest mini-ferments for sensory assessment and send samples to a commercial laboratory to quantify levels of smoke-derived volatiles in the wine. Significant novel research is being conducted using spectroscopic techniques coupled with machine learning modeling to assess grapevine smoke contamination and taint in grapes and wine, offering growers and winemakers additional tools to monitor grapevine smoke exposure and taint rapidly and non-destructively in grapes and wine.

Classification of Soybean Genotypes Assessed Under Different Water Availability and at Different Phenological Stages Using Leaf-Based Hyperspectral Reflectance

Monitoring of soybean genotypes is important because of intellectual property over seed technology, better management over seed genetics, and more efficient strategies for its agricultural production process. This paper aims at spectrally classifying soybean genotypes submitted to diverse water availability levels at different phenological stages using leaf-based hyperspectral reflectance. Leaf reflectance spectra were collected using a hyperspectral proximal sensor. Two experiments were conducted as field trials: one experiment was at Embrapa Soja in the 2016/2017, 2017/2018, and 2018/2019 cropping seasons, where ten soybean genotypes were grown under four water conditions; and another experiment was in the experimental farm of Unoeste University in the 2018/2019 cropping season, where nine soybean genotypes were evaluated. The spectral data collected was divided into nine spectral datasets, comprising single and multiple cropping seasons (from 2016 to 2019), and two contrasting crop-growing environments. Principal component analysis, applied as an indicator of the explained variance of the reflectance spectra among genotypes within each spectral dataset, explained over 94% of the spectral variance in the first three principal components. Linear discriminant analysis, used to obtain a model of classification of each reflectance spectra of soybean leaves into each soybean genotype, achieved accuracy between 61% and 100% in the calibration procedure and between 50% and 100% in the validation procedure. Misclassification was observed only between genotypes from the same genetic background. The results demonstrated the great potential of the spectral classification of soybean genotypes at leaf-scale, regardless of the phenological stages or water status to which plants were submitted.

Infrared Spectroscopy and Chemometric Applications for the Qualitative and Quantitative Investigation of Grapevine Organs

The fourth agricultural revolution is leading us into a time of using data science as a tool to implement precision viticulture. Infrared spectroscopy provides the means for rapid and large-scale data collection to achieve this goal. The non-invasive applications of infrared spectroscopy in grapevines are still in its infancy, but recent studies have reported its feasibility. This review examines near infrared and mid infrared spectroscopy for the qualitative and quantitative investigation of intact grapevine organs. Qualitative applications, with the focus on using spectral data for categorization purposes, is discussed. The quantitative applications discussed in this review focuses on the methods associated with carbohydrates, nitrogen, and amino acids, using both invasive and non-invasive means of sample measurement. Few studies have investigated the use of infrared spectroscopy for the direct measurement of intact, fresh, and unfrozen grapevine organs such as berries or leaves, and these studies are examined in depth. The chemometric procedures associated with qualitative and quantitative infrared techniques are discussed, followed by the critical evaluation of the future prospects that could be expected in the field.

Comparative quantification of chlorophyll and polyphenol levels in grapevine leaves sampled from different geographical locations

Near infrared spectroscopy (NIRS) and mid-infrared spectroscopy (MIRS) in combination with chemometric analysis were applied to discriminate the geographical origin of grapevine leaves belonging to the variety “Touriga Nacional” during different vegetative stages. Leaves were collected from plants of two different wine regions in Portugal (Dão and Douro) over the grapes maturation period. A sampling plan was designed in order to obtain the most variability within the vineyards taking into account variables such as: solar exposition, land inclination, altitude and soil properties, essentially. Principal component analysis (PCA) was used to extract relevant information from the spectral data and presented visible cluster trends. Results, both with NIRS and MIRS, demonstrate that it is possible to discriminate between the two geographical origins with an outstanding accuracy. Spectral patterns of grapevine leaves show significant differences during grape maturation period, with a special emphasis between the months of June and September. Additionally, the quantification of total chlorophyll and total polyphenol content from leaves spectra was attempted by both techniques. For this purpose, partial least squares (PLS) regression was employed. PLS models based on NIRS and MIRS, both demonstrate a statistically significant correlation for the total chlorophyll (R²_P = 0.92 and R²_P = 0.76, respectively). However, the PLS model for the total polyphenols, may only be considered as a screening method, because significant prediction errors, independently of resourcing on NIRS, MIRS or both techniques simultaneously, were obtained.

Eco-Friendly Estimation of Heavy Metal Contents in Grapevine Foliage Using In-Field Hyperspectral Data and Multivariate Analysis

Heavy metal monitoring in food-producing ecosystems can play an important role in human health safety. Since they are able to interfere with plants' physiochemical characteristics, which influence the optical properties of leaves, they can be measured by in-field spectroscopy. In this study, the predictive power of spectroscopic data is examined. Five treatments of heavy metal stress (Cu, Zn, Pb, Cr, and Cd) were applied to grapevine seedlings and hyperspectral data (350-2500 nm), and heavy metal contents were collected based on in-field and laboratory experiments. The partial least squares (PLS) method was used as a feature selection technique, and multiple linear regressions (MLR) and support vector machine (SVM) regression methods were applied for modelling purposes. Based on the PLS results, the wavelengths in the vicinity of 2431, 809, 489, and 616 nm; 2032, 883, 665, 564, 688, and 437 nm; 1865, 728, 692, 683, and 356 nm; 863, 2044, 415, 652, 713, and 1036 nm; and 1373, 631, 744, and 438 nm were found most sensitive for the estimation of Cu, Zn, Pb, Cr, and Cd contents in the grapevine leaves, respectively. Therefore, visible and red-edge regions were found most suitable for estimating heavy metal contents in the present study. Heavy metals played a significant role in reforming the spectral pattern of stressed grapevine compared to healthy samples, meaning that in the best structures of the SVM regression models, the concentrations of Cu, Zn, Pb, Cr, and Cd were estimated with R2 rates of 0.56, 0.85, 0.71, 0.80, and 0.86 in the testing set, respectively. The results confirm the efficiency of in-field spectroscopy in estimating heavy metals content in grapevine foliage.

Scenario-based discrimination of common grapevine varieties using in-field hyperspectral data in the western of Iran

Field spectroscopy is an accurate, rapid and nondestructive technique for monitoring of agricultural plant characteristics. Among these, identification of grapevine varieties is one of the most important factors in viticulture and wine industry. This study evaluated the discriminatory ability of field hyperspectral data and statistical techniques in case of five common grapevine varieties in the western of Iran. A total of 3000 spectral samples were acquired at leaf and canopy levels. Then, in order to identify the best approach, two types of hyperspectral data (wavelengths from 350 to 2500 nm and 32 spectral indices), two data reduction methods (PLSR and ANOVA-PCA) and two classification algorithms (LDA and SVM) were applied in a total of 16 scenarios. Results showed that the grapevine varieties were discriminated with overall accuracy of 89.88%-100% in test sets. Among the data reduction methods, the combination of ANOVA and PCA yielded higher performance as opposed to PLSR. Accordingly, optimal wavelengths in discrimination of studied grapevine varieties were located in vicinity of 695, 752, 1148, 1606 nm and 582, 687, 1154, 1927 nm at leaf and canopy levels, respectively. Optimal spectral indices were R680, WI, SGB and RATIO975_2, DattA, Greenness at leaf and canopy levels, respectively. Also, the importance of spectral regions in discriminating studied grapevine varieties was ranked as near-infrared > mid-infrared and red edge region > visible. As a general conclusion, the canopyspectral indices-ANOVA-PCA-SVM scenario discriminated the studied species most accurately.

Evaluation of a miniaturized NIR spectrometer for cultivar identification: The case of barley, chickpea and sorghum in Ethiopia

Crop cultivar identification is fundamental for agricultural research, industry and policies. This paper investigates the feasibility of using visible/near infrared hyperspectral data collected with a miniaturized NIR spectrometer to identify cultivars of barley, chickpea and sorghum in the context of Ethiopia. A total of 2650 grains of barley, chickpea and sorghum cultivars were scanned using the SCIO, a recently released miniaturized NIR spectrometer. The effects of data preprocessing techniques and choosing a machine learning algorithm on distinguishing cultivars are further evaluated. Predictive multiclass models of 24 barley cultivars, 19 chickpea cultivars and 10 sorghum cultivars delivered an accuracy of 89%, 96% and 87% on hold-out sample. The Support Vector Machine (SVM) and Partial least squares discriminant analysis (PLS-DA) algorithms consistently outperformed other algorithms. Several cultivars, believed to be widely adopted in Ethiopia, were identified with perfect accuracy. These results advance the discussion on cultivar identification survey methods by demonstrating that miniaturized NIR spectrometers represent a low-cost, rapid and viable tool. We further discuss the potential utility of the method for adoption surveys, field-scale agronomic studies, socio-economic impact assessments and value chain quality control. Finally, we provide a free tool for R to easily carry out crop cultivar identification and measure uncertainty based on spectral data.

Vineyard water status assessment using on-the-go thermal imaging and machine learning

The high impact of irrigation in crop quality and yield in grapevine makes the development of plant water status monitoring systems an essential issue in the context of sustainable viticulture. This study presents an on-the-go approach for the estimation of vineyard water status using thermal imaging and machine learning. The experiments were conducted during seven different weeks from July to September in season 2016. A thermal camera was embedded on an all-terrain vehicle moving at 5 km/h to take on-the-go thermal images of the vineyard canopy at 1.2 m of distance and 1.0 m from the ground. The two sides of the canopy were measured for the development of side-specific and global models. Stem water potential was acquired and used as reference method. Additionally, reference temperatures Tdry and Twet were determined for the calculation of two thermal indices: the crop water stress index (CWSI) and the Jones index (Ig). Prediction models were built with and without considering the reference temperatures as input of the training algorithms. When using the reference temperatures, the best models casted determination coefficients R2 of 0.61 and 0.58 for cross validation and prediction (RMSE values of 0.190 MPa and 0.204 MPa), respectively. Nevertheless, when the reference temperatures were not considered in the training of the models, their performance statistics responded in the same way, returning R2 values up to 0.62 and 0.65 for cross validation and prediction (RMSE values of 0.190 MPa and 0.184 MPa), respectively. The outcomes provided by the machine learning algorithms support the use of thermal imaging for fast, reliable estimation of a vineyard water status, even suppressing the necessity of supervised acquisition of reference temperatures. The new developed on-the-go method can be very useful in the grape and wine industry for assessing and mapping vineyard water status.

On-The-Go Hyperspectral Imaging Under Field Conditions and Machine Learning for the Classification of Grapevine Varieties

Grapevine varietal classification is an important plant phenotyping issue for grape growing and wine industry. This task has been achieved from destructive techniques like classic ampelography and DNA analysis under laboratory conditions. This work displays a new approach for the classification of a high number of grapevine (Vitis vinifera L.) varieties under field conditions using on-the-go hyperspectral imaging and different machine learning algorithms. On-the-go imaging was performed under natural illumination using a hyperspectral camera mounted on an all-terrain vehicle at 5 km/h. Spectra were acquired over two different leaf phenological stages on the canopy of 30 different varieties on a commercial vineyard located in La Rioja, Spain. A total of 1,200 spectral samples were generated. Support vector machines (SVM) and artificial neural networks (multilayer perceptrons, MLP) were used for the development of a large number of models, testing different algorithm parameters and spectral pre-processing techniques. Both classifiers yielded notable performance values and were able to train models with recall F1 scores and area under the receiver operating characteristic curve marks up to 0.99 for 5-fold cross validation. Statistical analyses supported that the best SVM kernel was linear and the best activation function for MLP was the hyperbolic tangent function. The prediction performance for individual varieties of MLP ranged from 0.94 to 0.99, displaying low levels of variability. In the case of SVM, slightly higher differences were obtained, ranging from 0.83 to 0.97 for individual varieties. These results support the possibility of deploying an on-the-go hyperspectral imaging system in the field capable of successfully classifying leaves from different grapevine varieties. This technology could thus be considered as a new useful non-destructive tool for plant phenotyping under field conditions.

Addressing Facts and Gaps in the Phenolics Chemistry of Winery By-Products

Grape and wine phenolics display a noticeable structural diversity, encompassing distinct compounds ranging from simple molecules to oligomers, as well as polymers usually designated as tannins. Since these compounds contribute critically to the organoleptic properties of wines, their analysis and quantification are of primordial importance for winery industry operators. Besides, the occurrence of these compounds has been also extensively described in winery residues, which have been pointed as a valuable source of bioactive phytochemicals presenting potential for the development of new added value products that could fit the current market demands. Therefore, the cumulative knowledge generated during the last decades has allowed the identification of the most promising compounds displaying interesting biological functions, as well as the chemical features responsible for the observed bioactivities. In this regard, the present review explores the scope of the existing knowledge, concerning the compounds found in these winery by-products, as well as the chemical features presumably responsible for the biological functions already identified. Moreover, the present work will hopefully pave the way for further actions to develop new powerful applications to these materials, thus, contributing to more sustainable valorization procedures and the development of newly obtained compounds with enhanced biological properties.

Data Mining and NIR Spectroscopy in Viticulture: Applications for Plant Phenotyping under Field Conditions

Plant phenotyping is a very important topic in agriculture. In this context, data mining strategies may be applied to agricultural data retrieved with new non-invasive devices, with the aim of yielding useful, reliable and objective information. This work presents some applications of machine learning algorithms along with in-field acquired NIR spectral data for plant phenotyping in viticulture, specifically for grapevine variety discrimination and assessment of plant water status. Support vector machine (SVM), rotation forests and M5 trees models were built using NIR spectra acquired in the field directly on the adaxial side of grapevine leaves, with a non-invasive portable spectrophotometer working in the spectral range between 1600 and 2400 nm. The ν-SVM algorithm was used for the training of a model for varietal classification. The classifiers’ performance for the 10 varieties reached, for cross- and external validations, the 88.7% and 92.5% marks, respectively. For water stress assessment, the models developed using the absorbance spectra of six varieties yielded the same determination coefficient for both cross- and external validations (R² = 0.84; RMSEs of 0.164 and 0.165 MPa, respectively). Furthermore, a variety-specific model trained only with samples of Tempranillo from two different vintages yielded R² = 0.76 and RMSE of 0.16 MPa for cross-validation and R² = 0.79, RMSE of 0.17 MPa for external validation. These results show the power of the combined use of data mining and non-invasive NIR sensing for in-field grapevine phenotyping and their usefulness for the wine industry and precision viticulture implementations.