Aplicación móvil de detección y clasificación de “la roya” en hojas de café robusta mediante aprendizaje automático

Manabí being a province endowed with natural resources and being agriculture one of the most important activities, seeks to provide optimized solutions with a view to agriculture and current technological trends. The purpose of this titling work was to develop a mobile application based on machine learning that would automate the detection and classification of “Roya” on robust coffee leaves. The images were obtained in the plots of the Agricultural Research and Innovation and Development City (CIIDEA) where robust coffee plantations were located, once images were obtained, the correct processing was carried out that involved determining when a leaf was infected and when not. The literature review was carried out to determine the most feasible machine-learning algorithm to be implemented. With this data, the mobile application is developed using the Extreme Programming development methodology, using machine-learning techniques, which allowed detecting this fungus in the robust coffee leaves. These results can help the areas that involve the study of this type of research, such as the project to characterize said disease in coffee crops, in addition, the development of the application and implementation of the algorithm provided a 97% detection effectiveness of "Roya".

• Main Author: Cusme Zambrano, Kevin Daniel (author)
• Other Authors: Loor Pinargote, Angélica María (author)
• Format: bachelorThesis
• Language: spa
• Published: 2019
• Subjects: Café robusta; Aprendizaje automático; Roya; Aplicación móvil
• Online Access: http://repositorio.espam.edu.ec/handle/42000/1104