RESUMO

The development of non-invasive methods and accessible tools for application to plant phenotyping is considered a breakthrough. This work presents the preliminary results using an electronic nose (E-Nose) and machine learning (ML) as affordable tools. An E-Nose is an electronic system used for smell global analysis, which emulates the human nose structure. The soybean (Glycine Max) was used to conduct this experiment under water stress. Commercial E-Nose was used, and a chamber was designed and built to conduct the measurement of the gas sample from the soybean. This experiment was conducted for 22 days, observing the stages of plant growth during this period. This chamber is embedded with relative humidity [RH (%)], temperature (°C), and CO2 concentration (ppm) sensors, as well as the natural light intensity, which was monitored. These systems allowed intermittent monitoring of each parameter to create a database. The soil used was the red-yellow dystrophic type and was covered to avoid evapotranspiration effects. The measurement with the electronic nose was done daily, during the morning and afternoon, and in two phenological situations of the plant (with the healthful soy irrigated with deionized water and underwater stress) until the growth V5 stage to obtain the plant gases emissions. Data mining techniques were used, through the software "Weka™" and the decision tree strategy. From the evaluation of the sensors database, a dynamic variation of plant respiration pattern was observed, with the two distinct behaviors observed in the morning (~9:30 am) and afternoon (3:30 pm). With the initial results obtained with the E-Nose signals and ML, it was possible to distinguish the two situations, i.e., the irrigated plant standard and underwater stress, the influence of the two periods of daylight, and influence of temporal variability of the weather. As a result of this investigation, a classifier was developed that, through a non-invasive analysis of gas samples, can accurately determine the absence of water in soybean plants with a rate of 94.4% accuracy. Future investigations should be carried out under controlled conditions that enable early detection of the stress level.

RESUMO

Repackaging and tampering with labels of foods to extend their shelf life is an illegal practice, increasingly common in some Brazilian coffee retail markets. Fast, easy-to-use, and low-cost analytical techniques for the large-scale screening of aging time have been demanded lately to fight the growth of these frauds in retail coffee markets. In this work, Fourier transform infrared spectroscopy was evaluated as a provider of relevant regressors, chemically explainable, aiming for predictive models for estimating the aging of roasted and packaged coffees during their shelf life. Spectra of two Coffea arabica varieties (Bourbon and Obatã) were periodically acquired during eleven months of storage. The most relevant absorption bands were selected, which showed a moderate correlation with the storage time. They were identified as responses from lipids, phenolic compounds, and carbohydrates. From those responsive bands, logistic regression (sigmoid functions) models were fitted for each coffee variety, as well as for both together. Predictive models for Bourbon and Obatã showed high performances in validation data, with r (Pearson correlation) above 0.92 and root mean square error (RMSE) below 43 days. For both varieties, the logistic model showed r greater than 0.83 and RMSE equal to 56 days. Results corroborate the methodological approach efficacy towards agile technological innovations in the coffee value chain, as well as opening new application fronts for estimating the aging of other foods.

Assuntos

RESUMO

One of the most important factors that interfere negatively in coffee global quality has been blends with defective beans, especially those called Black, Immature and Sour (BIS). The methods based on visual-manual estimation of defective beans have shown their inefficiency in coffee value chain for large-scale analysis. The lack of fast, accurate and robust analytical methods for BIS determination is still a research gap. Laser-Induced Breakdown Spectroscopy (LIBS) is a fast, low-cost and residue-free technique capable of performing multielemental determination and investigating organic composition of samples. In the present work, LIBS together with spectral processing and variable selection were evaluated to fit linear regression models for predicting BIS in blends. Models showed high capacity of prediction with RMSEP smaller than 3.8% and R2 higher than 80%. Most importantly, measurements are guided by chemical responses, which make LIBS-based methods less susceptible to the visual indistinguishability that occurs in manual inspections.

Assuntos

RESUMO

Huanglongbing (HLB) is the most recent and destructive bacterial disease of citrus and has no cure yet. A promising alternative to conventional methods is to use laser-induced breakdown spectroscopy (LIBS), a multi-elemental analytical technique, to identify the nutritional changes provoked by the disease to the citrus leaves and associate the mineral composition profile with its health status. The leaves were collected from adult citrus trees and identified by visual inspection as healthy, HLB-symptomatic, and HLB-asymptomatic. Laser-induced breakdown spectroscopy measurements were done in fresh leaves without sample preparation. Nutritional variations were evaluated using statistical tools, such as Student's t-test and analysis of variance applied to LIBS spectra, and the largest were found for Ca, Mg, and K. Considering the nutritional profile changes, a classifier induced by classification via regression combined with partial least squares regression was built resulting in an accuracy of 73% for distinguishing the three categories of leaves.

Assuntos

RESUMO

Huanglongbing (HLB) and citrus variegated chlorosis (CVC) are serious threats to citrus production and have caused considerable economic losses worldwide, especially in Brazil, which is one of the biggest citrus producers in the world. Neither disease has a cure nor an efficient means of control. They are also generally confused with each other in the field since they share similar initial symptoms, e.g., yellowing blotchy leaves. The most efficient tool for detecting these diseases is by polymerase chain reaction (PCR). However, PCR is expensive, is not high throughput, and is subject to cross reaction and contamination. In this report, a diagnostic method is proposed for detecting HLB and CVC diseases in leaves of sweet orange trees using attenuated total reflectance Fourier transform infrared spectroscopy and the induced classifier via partial least-squares regression. Four different leaf types were considered: healthy, CVC-symptomatic, HLB-symptomatic, and HLB-asymptomatic. The results show a success rate of 93.8% in correctly identifying these different leaf types. In order to understand which compounds are responsible for the spectral differences between the leaf types, samples of carbohydrates starch, sucrose, and glucose, flavonoids hesperidin and naringin, and coumarin umbelliferone were also analyzed. The concentration of these compounds in leaves may vary due to biotic stresses.

Assuntos

RESUMO

Laser induced breakdown spectroscopy (LIBS) is an atomic emission spectroscopy technique for simple, direct and clean analysis, with great application potential in environmental sustainability studies. In a single LIBS spectrum it is possible to obtain qualitative information on the sample composition. However, quantitative analysis requires a reliable model for analytical calibration. Multilayer perceptron (MLP), an artificial neural network, is a multivariate technique that is capable of learning to recognize features from examples. Therefore MLP can be used as a calibration model for analytical determinations. Accordingly, the present study proposes to evaluate the traditional linear fit and MLP models for LIBS calibration, in order to attain a quantitative multielemental method for contaminant determination in soil under sewage sludge application. Two sets of samples, both composed of two kinds of soils were used for calibration and validation, respectively. The analyte concentrations in these samples, used as reference, were determined by a reference analytical method using inductively coupled plasma optical emission spectrometry (ICP OES). The LIBS-MLP was compared to a LIBS-linear fit method. The values determined by LIBS-MLP showed lower prediction errors, correlation above 98% with values determined by ICP OES, higher accuracy and precision, lower limits of detection and great application potential in the analysis of different kinds of soils.

Assuntos