Precision Farming
R. Azadnia; A. Rajabipour; B. Jamshidi; M. Omid
Abstract
IntroductionApple is one of the most frequently consumed fruits in the world. It is a source of minerals, fiber, various biological compounds such as vitamin C, and phenolic compounds (natural antioxidants). The amount of nutrients plays a significant role in the growth, reproduction, and performance ...
Read More
IntroductionApple is one of the most frequently consumed fruits in the world. It is a source of minerals, fiber, various biological compounds such as vitamin C, and phenolic compounds (natural antioxidants). The amount of nutrients plays a significant role in the growth, reproduction, and performance of agricultural products and plants. Chemical inputs can be accurately managed by predicting these elements. Thus, timely and accurate monitoring and managing the status of crop nutrition is crucial for adjusting fertilization, increasing the yield, and improving the quality. This approach minimizes the application of chemical fertilizers and reduces the risk of environmental degradation. In crop plants, leaf samples are typically analyzed to diagnose nutrient deficiencies and imbalances, as well as to evaluate the effectiveness of the current nutrient management system. Therefore, the main aim of this study is to estimate the level of Nitrogen (N), Phosphorus (P), and Potassium (K) elements in the leaves of the apple tree using the non-destructive method of Visible/Near-infrared (Vis/NIR) spectroscopy at the wavelength range of 500 to 1000 nm coupled with chemometrics analysis.Materials and MethodsThis research investigated the potential of the Vis/NIR spectroscopy coupled with chemometrics analysis for predicting NPK nutrient levels of apple trees. In this study, 80 leaf samples of apple trees were randomly picked and transferred to the laboratory for spectral measurement. The Green-Wave spectrometer (StellarNet Inc, Florida, USA) was utilized to collect the spectral data. In the next step, the spectral data were transferred to the laptop using the Spectra Wiz software (StellarNet Inc, Florida, USA). For this purpose, spectroscopy of the leaf samples was done in interactance mode. Ten random points were selected on each leaf to capture reflectance spectra and the averaged spectrum was used to determine the reflectance (R). The data was then transformed into absorbance (log 1/R) for chemometrics analysis. Following the spectroscopy measurements, the NPK contents were measured using reference methods. Afterward, Partial Least Square (PLS) multivariate calibration models were developed based on the reference measurements and spectral information using different pre-processing techniques. To remove the unwanted effects, various pre-processing methods were utilized to obtain an accurate calibration model. To evaluate the proposed models, the Root Mean Square Error of calibration and prediction sets (RMSEC and RMSEP), as well as the correlation coefficient of calibration and prediction sets (rc and rp), and Residual Predictive Deviation (RPD) were calculated.Results and DiscussionThe statistical metrics were calculated for the evaluation of PLS models and the results indicated that the PLS models could efficiently predict the NPK contents with satisfactory accuracy. The model with the best performance for nitrogen prediction was based on the standard normal variate pre-processing method in combination with the second derivative (SNV+D2) and resulted in rc= 0.988, RMSEC=0.028%, rp=0.978, RMSEP=0.034%, and RPD of 7.47. The best model for P content prediction resulted in rc= 0.967, RMSEC=0.0051%, rp=0.958, RMSEP=0.0057%, and RPD of 5.96. Additionally, the PLS model based on MSC+D2 pre-processing method resulted in rc= 0.984, RMSEC=0.017%, rp=0.976, RMSEP=0.021%, and RPD of 7.10, indicating the high potential of PLSR model in predicting K content. Moreover, the weakest performing model was related to the estimation of P content without pre-processing with rc = 0.774, RMSEC = 0.013%, rp = 0.685, RMSEP = 0.018%, and RPD value of 1.87. Based on the obtained results, the proposed PLS models coupled with suitable pre-processing methods were able to predict the nutrient content with high precision.ConclusionField spectroscopy has recently gained popularity due to its portability, ease of use, and low cost. Consequently, the use of a portable system for estimating nutrient levels in the field can significantly save time and lower laboratory expenses. Therefore, due to the accuracy of the Vis/NIR spectroscopy technique and according to the obtained results, this method can be used to actualize a portable system based on Vis/NIR spectroscopy to estimate the nutrient elements needed by the apple trees in the orchards and to increase the productivity of the orchards.
A. Moghimi; A. Sazgarnia; M. H. Aghkhani
Abstract
IntroductionPistachio production has been adversely affected by Psylla, which is a devastating insect. The primary goal of this study was to select sensitive spectral bands to distinguish pistachio leaves infected by Psylla from healthy leaves. Diagnosis of psylla disease before the onset of visual cues ...
Read More
IntroductionPistachio production has been adversely affected by Psylla, which is a devastating insect. The primary goal of this study was to select sensitive spectral bands to distinguish pistachio leaves infected by Psylla from healthy leaves. Diagnosis of psylla disease before the onset of visual cues is crucial for making decisions about topical garden management. Since it is not possible to diagnose psylla disease even after the onset of symptoms with the help of color images by drones, hyperspectral and multispectral sensors are needed. The main purpose of this study was to extract spectral bands suitable for distinguishing healthy leaves from psylla leaves. For this purpose, in this paper, a new method for selecting sensitive spectral properties from hyperspectral data with the high spectral resolution is presented. The intelligent selection of sensitive bands is a convenient way to build multispectral sensors for a specific application (in this article, the diagnosis of psylla leaves). Knowledge of disease-sensitive wavelengths can also help researchers analyze multispectral and hyperspectral aerial images captured by satellites or drones.Materials and MethodsA total number of 160 healthy and diseased leaves were scanned in 64 spectral bands between 400-1100 nm with 10 nm spectral resolution. A random forest algorithm was used to identify the importance of features in classifying the dataset into diseased and healthy leaves. After computing the importance of the features, a clustering algorithm was developed to cluster the most important features into six clusters such that the center of clusters was 50 nm apart. To transfer the hyperspectral dataset into a multispectral dataset, the reflectance was averaged in spectral bands within ±15 nm of each cluster center and achieved six broad multispectral bands. Afterwards a support vector machine algorithm was utilized to classify the diseased and healthy leaves using both hyperspectral and multispectral datasets.Results and DiscussionThe center of clusters were 468 nm, 598 nm, 710 nm, 791 nm, 858 nm, and 1023 nm, which were calculated by taking the average of all the members assigned to the individual clusters. These are the most informative spectral bands to distinguish the pistachio leaves infected by Psylla from the healthy leaves. The F1-score was 90.91 when the hyperspectral dataset (all bands) was used, while the F1-score was 88.69 for the multispectral dataset. The subtle difference between the F1-scores indicates that the proposed pipeline in this study was able to select appropriately the sensitive bands while retaining all relevant information.ConclusionThe importance of spectral bands in the visible and near-infrared region (between 400 and 1100 nm) was obtained to identify pistachio tree leaves infected with psylla disease. Based on the importance of spectral properties and using a clustering algorithm, six wavelengths were obtained as the best wavelengths for classifying healthy and diseased pistachio leaves. Then, by averaging the wavelengths at a distance of 15 nm from these six centers, the hyperspectral data (64 bands) became multispectral (6 bands). Since the correlation between the wavelengths in the near-infrared region was very high (more than 95%), out of the three selected wavelengths in the near-infrared region (710, 791, and 1023), only the 710-nm wavelength, which was closer to the visible region, was selected. The results of classification of infected and diseased leaves using hyperspectral and multispectral data showed that the degree of classification accuracy decreases by about 2% and if only 4 bands are used, the degree of accuracy decreases by about 3%.The results of this study revealed that the proposed framework could be used for selecting the most informative spectral bands and accordingly develop custom-designed multispectral sensors for disease detection in pistachio. In addition, we could reduce the dimensionality of the hyperspectral datasets and avoid the issues related to the curse of dimensionalitylity.
