Detección de deficiencias nutricionales en el maíz mediante inteligencia artificial en la finca dos hermanos.

Corn is a fundamental cereal for the diets of various countries, where nutrient deficiencies can cause damage to both grain yield and biomass. The objective of this study, conducted at the "Dos Hermanos" farm, was to classify nutrient deficiencies in corn leaves using artificial intelligence, specifically for the three primary nutrients essential for healthy growth: nitrogen, phosphorus, and potassium, as well as to identify healthy leaves. For this purpose, a dataset was created with images of corn leaves displaying visual characteristics according to the nutrient deficiencies present or if they were healthy. These data were preprocessed to then extract their features using the original pixel values and a Gabor filter bank. A classification model was then created using Support Vector Machine (SVM), Random Forest (RF), and k-Nearest Neighbors (k-NN). Additionally, GridSearchCV was employed to compare model performance with and without cross-validation. According to the evaluated metrics, the model with the best performance was Support Vector Machine (SVM), achieving 1.0% in all metrics without cross-validation and a 0.98% best score when using optimal parameters determined by GridSearchCV.

• Päätekijä: San Pedro Cevallos, Linley Liliana Torres Quijije, Angel Iván (author)
• Aineistotyyppi: bachelorThesis
• Kieli: spa
• Julkaistu: 2025
• Aiheet: Kernel; Hiperparámetros; Python; Codificación; Etiquetas
• Linkit: https://repositorio.uteq.edu.ec/handle/43000/9119