Coffee crop detection by automatic classification using spectral and textural attributes and illumination factor

Coffee, an important product in Brazilian exports, needs constant monitoring, so that systems and forecasting of existing crops can be reliable. Orbital imagery of medium spatial resolution are tools with great potential for land use mapping and identification of agricultural crops. This research ev...

Descripción completa

Detalles Bibliográficos
Autores: Marujo, Rennan de Freitas Bezerra, Moreira, Maurício Alves, Volpato, Margarete Marin Lordelo, Alves, Helena Maria Ramos
Tipo de recurso: artículo
Estado:Versión publicada
Fecha de publicación:2017
País:Brasil
Institución:Universidade Federal de Lavras (UFLA)
Repositorio:Coffee Science (Online)
Idioma:portugués
OAI Identifier:oai:coffeescience.ufla.br:article/1176
Acceso en línea:https://coffeescience.ufla.br/index.php/Coffeescience/article/view/1176
Access Level:acceso abierto
Palabra clave:Remote sensing
SRTM
Landsat-8
Sensoriamento remoto
Descripción
Sumario:Coffee, an important product in Brazilian exports, needs constant monitoring, so that systems and forecasting of existing crops can be reliable. Orbital imagery of medium spatial resolution are tools with great potential for land use mapping and identification of agricultural crops. This research evaluated the performance of the object based classification, applied in OLI/Landsat-8 images, with the purpose of mapping of coffee crops. Three analyzes were made, the first one using exclusively spectral attribute, the second one including textural attributes and the third also considering illumination classes. Six OLI/Landsat-8 multispectral images were used, representing three different coffee phenological stages: fructification, graining and rest. The validation of the classifications was performed by the Monte Carlo method using reference images obtained by visual interpretation. The classification using exclusively spectral attributes resulted an accuracy of 57% for coffee class. There was no phenological stage that provided greater accuracy to the coffee class in the automatic classification of OLI/Landsat-8 images. The results demonstrate that texture is important for coffee detection, thus visual interpretation remains an important step to minimize classification errors.