Creating a normalized difference vegetation index (NDVI)
Our first example will be a normalized difference vegetation index (NDVI). NDVIs are used to show the relative health of plants in an area of interest. An NDVI algorithm uses satellite or aerial imagery to show relative health by highlighting the chlorophyll density in plants. NDVIs use only the red and near-infrared bands. The formula of an NDVI is as follows, which we’ll refer to as Eq (1):
NDVI = (Infrared – Red) / (Infrared + Red)
The goal of this analysis is to produce, to begin with, a multispectral image containing infrared and red bands, and end up with a pseudo color image using seven classes, which color the healthier plants darker green, less-healthy plants lighter green, and bare soil brown.
Because the health index is relative, it is important to localize the area of interest. You could perform a relative index for the entire globe, but vast areas such as the Sahara desert, on the low-vegetation...
