| Popis: |
The objective of this work was to use digital images from a balloon platform for identifying nitrogen status of a pasture. Five levels of N were applied to a pasture field of Brachiaria decumbens in a completely random design with six replications. Images were acquired using a pair of cameras (CMOS color camera and a monochrome camera with optical filter that block visible light) mounted in a balloon platform. The images were acquired at heights of 15, 20, 25 and 30 meters on 28 days after N application (DAA). Portable chlorophyll measurement (SPAD), leaf N content and dry matter yield of the grass were done on 32 DAA. The NDVI, Green-NDVI and SAVI vegetation indices were evaluated to detect N grass differences. It was possible to fit linear model of first or second grade for all tested variables. The green-NDVI presented higher correlations with SPAD, leaf N content and dry matter than the others indices for all heights. The correlation between the indices and SPAD, leaf N content and dry matter increased with the balloon height. |
