Desarrollo de un sistema de visión por computadora con Raspberry Pi para la detección temprana de la Moniliasis y la Fitóftora en el cultivo de cacao criollo de la finca Villa Laura en el Huila

Through this research, a computer vision system was developed with Raspberry Pi to contribute to the early detection of Moniliasis and Phytophthora, which currently affect cocoa the most, and from there, generate alternatives to minimize their impact. Neural networks were used for this study, in order to detect diseases in cocoa, the detection model used is the SSD MobileNet V2, an architecture used for object detection, trained and evaluated by a set of o wn images. The images come from the Villa Laura farm where a dataset of 540 images was extracted, the neural network, once trained, calculates the performance metrics, obtaining as a result an average accuracy of 0.783, and an average recall of 0.85. These results are very favorable for the model entered.

• Autores: Andrade Becerra, Cristian David
• Tipo de recurso: Trabajo de grado de pregrado
• Fecha de publicación: 2022
• Institución: Universidad Antonio Nariño
• Repositorio: Repositorio UAN
• Idioma: spa
• OAI Identifier: oai:repositorio.uan.edu.co:123456789/7905
• Acceso en línea: http://repositorio.uan.edu.co/handle/123456789/7905
• Palabra clave: Cacao; Redes Neuronales; Raspberry; enfermedades; Cocoa; Neural Networks; Raspberry; diseases
• Rights: openAccess
• License: Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)