Bibliographic citations
Sangama, G., (2023). Reconocimiento de patrones visuales para detección de enfermedades en plantaciones de arroz del sector Mishquiyacu - Morales [Universidad Nacional de San Martín. Fondo Editorial]. http://hdl.handle.net/11458/5750
Sangama, G., Reconocimiento de patrones visuales para detección de enfermedades en plantaciones de arroz del sector Mishquiyacu - Morales []. PE: Universidad Nacional de San Martín. Fondo Editorial; 2023. http://hdl.handle.net/11458/5750
@misc{renati/1056077,
title = "Reconocimiento de patrones visuales para detección de enfermedades en plantaciones de arroz del sector Mishquiyacu - Morales",
author = "Sangama Vela, Gabriela Abigail",
publisher = "Universidad Nacional de San Martín. Fondo Editorial",
year = "2023"
}
Early detection of diseases in rice plantations is crucial to ensure crop health and agricultural productivity. In this study, an approach based on a convolutional neural network and visual pattern recognition is proposed for the automatic detection of the most common diseases in rice plantations. A dataset consisting of 1874 rice leaf images was used, classified into three categories: rice blast, brown spot, and healthy. For model training and evaluation, an open-access dataset was used in combination with images taken by a smartphone in the Mishquiyacu sector, located in the Morales district. The dataset was divided into 70% for training, 20% for validation, and 10% for testing. The InceptionV3 convolutional neural network architecture was employed, known for its ability to extract relevant features from images, applying transfer learning to connect the pretrained model to a classification layer with 3 categories. Additionally, data preprocessing and augmentation techniques were applied to improve the model's generalization. The results obtained showed high accuracy in disease detection in rice plantations. The model achieved an accuracy of 96.79% on the test set, surpassing the AlexNet and ResNet50 models. This result was further evaluated using evaluation metrics such as precision, accuracy, F1 score, recall, Cohen's kappa, and area under the ROC curve, demonstrating the effectiveness of the model in classifying the provided data.
This item is licensed under a Creative Commons License