[ID 172] CORN PRODUCTIVITY PREDICTION IN THE CERRADO REGION USING SATELLITE IMAGES

Abstract

INTRODUCTION: Currently, there is a growing demand for modernization with regard to food production in the world. Precision Agriculture has been standing out in this scenario for offering useful information so that a high rate of production and added value can be achieved, especially in this project, the production of corn. OBJECTIVE: In this work, the objective was to evaluate the temporal behavior of the NDVI and its relationship with the productivity of the maize crop in order to generate productivity prediction models in plant populations and create a crop prediction model for the maize crop. MATERIAL AND METHODS: An image was acquired from the MSI-Sentinel 2 sensor, from the mapping of slope forms (Topodata) to generate vegetation indices, the temporal relationship of these indices with productive parameters of the culture was made. Manual analysis was performed at 20 points within the area. The analysis of results was made from the statistical description and literary basis. Only open software was used. RESULTS: Using regression analysis, the correlation of NDVI, manual average and the productivity obtained was compared, in addition, the variation of NDVI values ​​around a 15-meter area was compared to compare how variable was the result of the worked area. In the project. CONCLUSION: It was possible to observe that there was no correlation, however, analyzing each of the sampling points and the variation of the NDVI values ​​around in a radius of 15 meters, another result was found, the greater the variation, the greater the index of productivity.