Filters
Advancements in plant protection - the application of machine learning to the detection of maize infestations
Plant infestations cause significant economic losses in agriculture, necessitating rapid and accurate detection for optimized agrotechnical operations and reduced environmental pollution. This study addresses this challenge by proposing a customized convolutional neural network (CNN) architecture for detecting corn leaf worm infestations in maize. The research focuses on developing unique CNN models through extensive experimentation, systematically adjusting hyperparameters like optimizers, filter numbers, and kernel sizes. The study’s main contributions include the design of an accurate CNN classifier, and its implementation in a user-friendly smartphone application. The research highlights the importance of hyperparameter tuning in CNN performance, demonstrating that optimal configurations lead to high accuracy (up to 95% for accuracy, precision, recall, specificity, and F1-score). While the current model focuses on a single pest, the findings underscore the potential of custom CNN classifiers in vision systems for automated crop inspection, offering a promising solution for minimizing crop losses and the environmental impact of chemical plant protection products.
Digital Innovations in Agriculture
Digital agriculture, defined as the analysis and collection of various farm data, is constantly evolving [...]
Predicting Starch Content in Early Potato Varieties Using Neural Networks and Regression Models: A Comparative Study
Starch content serves as a crucial indicator of the quality and palatability of potato tubers. It has become a common practice to evaluate the polysaccharide content directly in tubers freshly harvested from the field. This study aims to develop models that can predict starch content prior to the harvesting of potato tubers. Very early potato varieties were cultivated in the northern and northwestern regions of Poland. The research involved constructing multiple linear regression (MLR) and artificial neural network (ANN-MLP) models, drawing on data from eight years of field trials. The independent variables included factors such as sunshine duration, average daily air temperatures, precipitation, soil nutrient levels, and phytophenological data. The NSM demonstrated a higher accuracy in predicting the dependent variable compared to the RSM, with MAPE errors of 7.258% and 9.825%, respectively. This study confirms that artificial neural networks are an effective tool for predicting starch content in very early potato varieties, making them valuable for monitoring potato quality.
Insights into Drought Tolerance of Tetraploid Wheat Genotypes in the Germination Stage Using Machine Learning Algorithms
Throughout germination, which represents the initial and crucial phase of the wheat life cycle, the plant is notably susceptible to the adverse effects of drought. The identification and selection of genotypes exhibiting heightened drought tolerance stand as pivotal strategies aimed at mitigating these effects. For the stated objective, this study sought to evaluate the responses of distinct wheat genotypes to diverse levels of drought stress encountered during the germination stage. The induction of drought stress was achieved using polyethylene glycol at varying concentrations, and the assessment was conducted through the application of multivariate analysis and machine learning algorithms. Statistical significance (p < 0.01) was observed in the differences among genotypes, stress levels, and their interaction. The ranking of genotypes based on tolerance indicators was evident through a principal component analysis and biplot graphs utilizing germination traits and stress tolerance indices. The drought responses of wheat genotypes were modeled using germination data. Predictions were then generated using four distinct machine learning techniques. An evaluation based on R-square, mean square error, and mean absolute deviation metrics indicated the superior performance of the elastic-net model in estimating germination speed, germination power, and water absorption capacity. Additionally, in assessing the criterion metrics, it was determined that the Gaussian processes classifier exhibited a better performance in estimating root length, while the extreme gradient boosting model demonstrated superior performance in estimating shoot length, fresh weight, and dry weight. The study’s findings underscore that drought tolerance, susceptibility levels, and parameter estimation for durum wheat and similar plants can be reliably and efficiently determined through the applied methods and analyses, offering a fast and cost-effective approach.
Compact Convolutional Transformer (CCT)-Based Approach for Whitefly Attack Detection in Cotton Crops
Cotton is one of the world’s most economically significant agricultural products; however, it is susceptible to numerous pest and virus attacks during the growing season. Pests (whitefly) can significantly affect a cotton crop, but timely disease detection can help pest control. Deep learning models are best suited for plant disease classification. However, data scarcity remains a critical bottleneck for rapidly growing computer vision applications. Several deep learning models have demonstrated remarkable results in disease classification. However, these models have been trained on small datasets that are not reliable due to model generalization issues. In this study, we first developed a dataset on whitefly attacked leaves containing 5135 images that are divided into two main classes, namely, (i) healthy and (ii) unhealthy. Subsequently, we proposed a Compact Convolutional Transformer (CCT)-based approach to classify the image dataset. Experimental results demonstrate the proposed CCT-based approach’s effectiveness compared to the state-of-the-art approaches. Our proposed model achieved an accuracy of 97.2%, whereas Mobile Net, ResNet152v2, and VGG-16 achieved accuracies of 95%, 92%, and 90%, respectively.
Exploring Digital Innovations in Agriculture: A Pathway to Sustainable Food Production and Resource Management
Today’s agriculture faces numerous challenges due to climate change, a growing population and the need to increase food productivity [...]
Neural Image Analysis for the Determination of Total and Volatile Solids in a Composted Sewage Sludge and Maize Straw Mixture
Waste management is one of most important challenges in environmental protection. Much effort is put into the development of waste treatment methods for further use. A serious problem is the treatment of municipal sewage sludge. One method that is useful for this substrate is composting. However, it is reasonable to compost a sewage sludge mixed with other substrates, such as maize straw. To carry out the composting process properly, it is necessary to control some parameters, including the total solids and volatile solids content in the composted mixture. In this paper, a method for the determination of the total solids and volatile solids content based on image analysis and neural networks was proposed. Image analysis was used for the determination of the colour and texture parameters. The three additional features describing the composted material were percentage of sewage sludge, type of maize straw, and stage of compost maturity. The neural models were developed based on various combinations of the input parameters. For both the total solids and volatile solids content, the most accurate models were obtained using all input parameters, including 30 parameters for image colour and texture and three features describing the composted material. The uncertainties of the developed models, expressed by the MAPE error, were 2.88% and 0.59%, respectively, for the prediction of the total solids and volatile solids content.
Combining Machine Learning Algorithms with Earth Observations for Crop Monitoring and Management
Combining machine learning algorithms with Earth observations has great potential in the context of crop monitoring and management, which is essential in the face of global challenges related to food security and climate change [...]
Quantifying Nutrient Content in the Leaves of Cowpea Using Remote Sensing
Although hyperspectral remote sensing techniques have increasingly been used in the nutritional quantification of plants, it is important to understand whether the method shows a satisfactory response during the various phenological stages of the crop. The aim of this study was to quantify the levels of phosphorus (P), potassium (K), calcium (Ca) and zinc (Zn) in the leaves of Vigna Unguiculata (L.) Walp using spectral data obtained by a spectroradiometer. A randomised block design was used, with three treatments and twenty-five replications. The crop was evaluated at three growth stages: V4, R6 and R9. Single-band models were fitted using simple correlations. For the band ratio models, the wavelengths were selected by 2D correlation. For the models using partial least squares regression (PLSR), the stepwise method was used. The model showing the best fit was used to estimate the phosphorus content in the single-band (R² = 0.62; RMSE = 0.54 and RPD = 1.61), band ratio (R² = 0.66; RMSE = 0.65 and RPD = 1.52) and PLSR models, using data from each of the phenological stages (R² = 0.80; RMSE = 0.47 and RPD = 1.66). Accuracy in modelling leaf nutrients depends on the phenological stage, as well as the amount of data used, and is more accurate with a larger number of samples.