Mapping soya bean and corn crops in the State of Paraná, Brazil, using EVI images from the MODIS sensor
Mapping soya bean and corn crops in the State of Paraná, Brazil, using EVI images from the MODIS sensor
Denise Maria Grzegozewski, Jerry Adriani Johann, Miguel Angel Uribe-Opazo,Erivelto Mercante & Alexandre Camargo Coutinho
http://dx.doi.org/10.1080/01431161.2016.1148285
ABSTRACT: This study aimed to map, separate, and estimate soya bean andcorn crop areas in Paraná State, Brazil, in the harvest years 2012/13and 2013/14, using the enhanced vegetation index (EVI) imagesfrom the Moderate Resolution Imaging Spectroradiometer(MODIS) sensor. Thus, two methodologies were integrated, thefirst considering heterogeneity on the dates of crop cycles, thescenes required to generate images of minimum and maximumvegetation indexes, creating a colour composite red, green, andblue (RGB), and identifying two cultures simultaneously. In thesecond methodology, soya bean and corn were identified andmapped using the selection of pure pixels and the supervisedclassification algorithm Spectral Angle Mapper (SAM). In order toavoid overlapping areas, we multiplied the results from the firstand second methodologies to obtain the final separation. The finalvalidation of the mapping was compared to official data, identifyinghigh correlation to crops. Based on Medium-Resolution LinearImaging Self-Scanner (LISS-III) and Land Remote Sensing Satellite(Landsat-8) images, the similarity of global accuracy (GA) andkappa accuracy indices was determined, being classified as goodand excellent, respectively. It showed that the use of the twoconsortium methodologies for separation and overlap elimination of these crops in the state of Paraná was efficient.
KEYWORDS: Annual agriculture; timeseries; vegetation index
Acessos: 5125