For this post, let’s focus specifically on how to leverage remote sensing images in Agriculture(to monitor plant health). Before jumping into the aerial images, let’s take a moment to explore some of the sources from where these images could be retrieved. As we all know, multiple earth observation satellites take pictures of the earth as they orbit around. These images carry a wide range of information that could be applied to many areas and one such field is Agriculture(precision farming).
Let’s consider some of the popular aerial Imagery Data Sources including images from Landsat-8 and Sentinel-2. Landsat-8 is the 8th in the series of Landsat satellites launched and is the latest one in the row. It comprises two sensors (1) Operational Land Imager(OLI) (2) Thermal Infrared Sensor(TIRS) whose main job is to capture images of earth in various spectral bands. OLI captures 9 bands(visible, infrared etc) whereas the TIRS sensor gathers information related to 2 thermal bands. The complete breakdown of the bands could be found on the USGS website.
Notes: Spectral bands are a range of wavelengths in the electromagnetic spectrum. And every range is referred to as bands. Some of the bands are called ‘reflective bands’ as they depend on the amount of light reflected back from the earth’s surface.
Sentinel-2 is another earth observation satellite, a product of the Copernicus Programme monitored by the European Space Agency. It has 13 spectral channels covering visible, near-infrared and short wave infrared wavelengths.
Now we know that we have satellite images available in a wide range of bands(wavelengths). The next natural question would be, which one to choose while dealing with vegetation. Every wavelength has its own unique characteristics capable of reflecting some details better than others. Instead of confining to one particular band, we can combine multiple wavelengths(more pronounced for vegetation) and create indices out of the mix. Let’s discuss some of those indices,
Normalized Difference Vegetation Index: The most widely used index to distinguish the rich green vegetation from the dull area. The formula for NDVI is (NIR — Red) / (NIR + Red) and the value ranges from -1 to 1. If we are using Sentinel-2 images, then the index = (Band8 — Band4)/(Band8+Band4) It is a dimensionless quantity as the numerator and denominator units cancel each other. The minimum value -1 indicates the presence of water, whereas smaller values such as 0.1, -0.1 denote barren lands and for green vegetation, the value will be higher(more positive).
But why NDVI is capable of capturing the details better?
“The pigment in plant leaves, chlorophyll, strongly absorbs visible light (from 0.4 to 0.7 µm) for use in photosynthesis. The cell structure of the leaves, on the other hand, strongly reflects near-infrared light (from 0.7 to 1.1 µm). The more leaves a plant has, the more these wavelengths of light are affected, respectively.” — from wiki
Even though NDVI is a suitable choice to obtain the vegetation density, it fails in a few scenarios. For instance, it is tough to distinguish between the medium dense area from super-high regions. It is also more susceptible to noise in the data(from NIR and R bands) and hence not Robust.
The other indices include the Leaf Area Index(LAI), Soil Adjusted Vegetation Index(SAVI) and Enhanced Vegetation Index(EVI) which also carry meaningful information in regard to the green regions.
NDVI - You Don't Always Need Deep Learning for Satellite Imagery
While looking for an ML solution to understand the relationship between climate change and forced displacement in…
15 Free Satellite Imagery Data Sources - GIS Geography
The secret is there are dozens of authoritative sources with high-quality satellite imagery. But wait there's more. Not…