TexVeg: The Next-Generation Vegetation Index

Vegetation mapping, monitoring and characterization are some of the most important and prolific exercises that employ satellite image processing in the agriculture and environmental monitoring sectors. Scientists can determine the presence, health and vigor, and seasonal variations of vegetated features from space.

Many vegetation indexes have been developed. By far, the most popular is Normalized Difference Vegetation Index (NDVI). Variations of this index concentrate on reducing background noise due to influences from atmosphere or soil. Another version, Normalized Difference Red Edge index (NDRE), was created to monitor crops or trees later in the growing season, particularly when plants have low chlorophyl content.¹

There are limits to how well vegetation can be characterized from satellite imagery. These indexes are great at understanding if vegetation exists at a location and if the vegetated features are healthy-but they are not as effective at understanding the stage of growth, height or density of plants reliably from a single high-resolution image.

I developed an algorithmic measurement based on plant texture and chlorophyll content. Texture is useful for determining difference in height among features in imagery; Maxar's sophisticated texture measurement is tuned particularly well to plants. By itself, the texture measurement cannot be used to identify vegetated versus nonvegetated features. And by itself, a vegetation index cannot tell the difference between short and tall vegetation. Used together in a novel way, Maxar's texture measurement and a vegetation index can identify vegetative features and determine the density and height of the vegetation- we call that result TexVeg. This is a great way to monitor vegetation for growth, which is beneficial to the energy and agriculture industries for a variety of use cases that I will outline in this blog post.