Clustering analysis applied to NDVI/NOAA multitemporal images to improve the monitoring process of sugarcane crops

This paper discusses how to take advantage of clustering techniques to analyze and extract useful information from multi-temporal images of low spatial resolution satellites to monitor the sugarcane expansion. Additionally, we introduce the SatImagExplorer system that was developed to automatically extract time series from a huge volume of remote sensing images as well as provide algorithms of clustering analysis and geospatial visualization. According to experiments accomplished with spectral images of sugarcane fields, this proposed approach can be satisfactorily used in crop monitoring.