La titularidad de los derechos patrimoniales de esta obra pertenece a las instituciones editoras. Su uso se rige por una licencia Creative Commons BY-NC-SA 4.0 Internacional, https://creativecommons.org/licenses/by-nc-sa/4.0/legalcode.es, para un uso diferente consultar al responsable jurídico del repositorio por medio del correo electrónico gabriel.ascanio@icat.unam.mx.
Ver términos de la licenciaKokate, Jayesh Krishnarao, et al. (2023). Deep Learning Based Leaf Disease Classification. Journal of Applied Research and Technology; Vol. 21 Núm. 5, 2023; 764-771. Recuperado de https://repositorio.unam.mx/contenidos/4148923
Autor(es)
Kokate, Jayesh Krishnarao; Kumar, Sunil; Kulkarni, Anant G
Tipo
Artículo de Investigación
Área del conocimiento
Ingenierías
Título
Deep Learning Based Leaf Disease Classification
Fecha
2023-10-30
Resumen
A plant in its healthy state can produce its crops to the utmost of its genetically defined potential. However, on the other hand, a plant which is infected by an infection causing agent which directly or indirectly interferes with the plants’ growth or its functioning. A disease may interfere several processes in a plant’s metabolism. ; Continuous manual monitoring of farm and leaves of the trees by an expert is not possible, as it would be very expensive and time-consuming. Identification of correct disease accurately when it first appears on the plant is a decisive footstep for proper managing and diseases control in fields. Thus, an automated way of identifying the diseases and accurate classification of disease will play an important role in taking appropriate action for stopping the further crop and yield damage. This paper presents a Xception model of deep learning for identification and classification of the leaf diseases, average accuracy of 98% is resulted in this work. Xception model is utilised in this work and results are presented in terms of precision, recall, and f1-score for the 11 disease classes of tomato leaf.
Tema
Cnn, Plant Diseases, Computer Vision, Classification
Idioma
eng
ISSN
ISSN electrónico: 2448-6736; ISSN: 1665-6423
