The use of machine learning in digital processing of satellite images applied to coffee crop.

Abstract

Remote sensing can be used to monitor and estimate, with reasonable correct answers, the yield, plant health, and coffee nutrition. Satellite-coupled sensors can obtain information about the spectral signature of the crop, on a time scale, in order to monitor and detect phenological changes. However, the accumulation of data obtained by orbital sensors makes it difficult to understand the relationship between the aspects of coffee. Thus, machine learning can perform data mining and meet the spectral signature patterns that constitute coffee behavior. This literature review sought the survey of research that used machine learning tools applied in digital image processing from satellites for coffee crop monitoring.