Relation of Normalized Difference Vegetation Index with Some Spectral Bands of Satellite Images

The satellite image is the current graphic document, containing the real data of the terrestrial surface when it was acquired, based on which the soil and the vegetation cover can be evaluated. The NDVI index is used for a large number of the vegetation properties. In this research we analyzed the variation of NDVI index depending on R, G, B and NIR spectral bands, based on the LANDSAT 8 images. The relation between NDVI index with the respective spectral bands was describe most accurately by an exponential regression equations of curves with negative coefficient in the case of R, G, B spectrum and by a logarithmic function for the NIR spectrum. The results show statistical certainty for all the relations of NDVI index with the analyzed spectral bands (RED: NDVI p<<0.001; r = -0.711), (GREEN: NDVI p<<0.001; r = -0.637), (BLUE: NDVI p<<0.001; r = -0.644), (NIR: NDVI p<<0.001; r = 0.734). The relations of NDVI index with others synthetic indices was analyzed according with some linear regression function with statistical certainty of the results; relation NDVI: NDBR (p<<0.001; r = 0.952); relation NDVI: NDWI (p<<0.001; r = 0.977); relation NDVI: NDMI (p<<0.001; r = 0.869), the confidence interval of 0.95.