Clasificacíon de variedades de planta de pitayo usando imágenes multiespectrales Clasificacíon de pitayo usando imágenes multiespectrales

Abstract

This study addresses the classification of Stenocereus queretaroensis (pitayo) varieties according to fruit color (red, yellow, and orange) using multispectral images obtained with a DJI Mavic 3M drone. A dataset with spectral features was built to train five machine learning models (KNN, SVM, Decision Tree, Random Forest, and Logistic Regression). The best results were achieved with the Decision Tree and Logistic Regression models, reaching an accuracy of 70 %, demonstrating that the integration of multispectral imagery, vegetation index (NDVI), and artificial intelligence repre