Comparación de índices de vegetación en base a imágenes multiespectrales en el cultivo de maíz con tres densidades de siembra

Abstract

The cultivation of corn is important at a global, national and regional level, it is part of food security. Crop management at a specific site level requires complicated tasks, availability of economic resources and labor. Currently, technologies are being adopted to monitor crops faster and optimize resources, such as the use of environmental remote sensing, specifically using multispectral images and vegetation indices. Given the importance of the crop and the feasibility of using vegetation indices, the project aimed to compare four vegetation indices: (NDVI, GRRI, GRVI, ARVI) on multispectral images of corn cultivation with three planting densities to evaluate their behavior. High spatial resolution images obtained by RPAS in two flights were used, with spatial resolution between 6 and 7 cm/pixel and flight height of 80 m. The orthophotomosaics were obtained by post-processing in the PIX4D software, the indices using the QGIS software and the data were processed in R-Studio. Correlations, linear regressions, validation of the models and comparison of the RMSE were carried out. As a result, 2 multispectral and 2 RGB orthophotos were obtained. 4 vegetation indices NDVI, GRRI, GRVI, ARVI were carried out for the 2 flights at 81 and 133 dds. The vegetation index with the highest precision according to the RMSE comparison is the NDVI (D1= 3.65 and 3.14), (D3=3.41 and 2.85) in the two flights respectively. It is concluded that the NDVI is the index with the greatest precision and can be used to monitor corn with planting distances of 25x80 cm and 75x 80 cm, at 81 and 133 dds.