Comparing Phenometrics Extracted From Dense Landsat-Like Image Time Series for Crop Classification

• Main Authors: Nascimento Bendini, Hugo do, Maria Garcia Fonseca, Leila, Schwieder, Marcel, Korting, Thales Sehn, Rufin, Philippe, Arco Sanches, Ieda Del, Leitao, Pedro J., Hostert, Patrick
• Format: Conference Proceeding
• Language: English
• Subjects: Agriculture; Artificial satellites; Big data; Earth; land use and land cover mapping; Measurement; multi-sensor; Radio frequency; random forest; Remote sensing; Time series analysis
• ISSN: 2153-7003
• DOI: 10.1109/IGARSS.2019.8898139

In this research, we compared two different sets of land surface phenological metrics (phenometrics) derived from dense satellite image time series to classify agricultural land in the Cerrado biome. We derived phenometrics from a dense Enhanced Vegetation Index (EVI) data cube with an 8-day temporal resolution and subjected them to classification using the Random Forest (RF) algorithm. We used a hierarchical classification with four levels, from land cover to crop rotation classes. We then evaluated the classification results comparing the use of phenometrics extracted using TIMESAT software [1], those obtained by polar representation, proposed by Körting et al. (2013) and the combination of both. We concluded that the accuracies of semi-perennial and winter crop classes increase substantially when using TIMESAT metrics combined with Polar features, and the misclassifications between single crops with non commercial crops are reduced.