MAPPING COFFEE AGRONOMIC PARAMETERS BASED ON REMOTELY PILOTED AIRCRAFT IMAGES

Abstract

The cultivation of coffee beans has increasingly been shown to be one of the predominant crops in the current market. With this, it’s necessary to understand and monitor the variability existing in large crops around the world. Therefore, remote sensing techniques aimed at monitoring the spectral variability of a crop makes it possible to map and study spatial variability and production limiting factors. Among several variables, this work aimed to map parameters related to productivity, such as chlorophyll content and leaf area of a coffee crop located near the municipality of Monte Carmelo - MG. The mapping was generated from regression models through the relationship between samples obtained in situ with the radiometric value of images taken by a remotely piloted aircraft at 70 and 120 meters high. The results showed that for higher flight the correlation between field measurements and radiometry of images was better. The accuracy of the estimator models showed a better correlation with the TGI indices (r = 0.536 and RMSE = 16.43%) for chlorophyll and NDVI (r = 0.484 and RMSE = 15.87%) to the leaf area.