MSU logo Minority Serving Institution Partnership Development Program

Remote Sensing: Normalized Difference Vegetation Index

The Normalized Difference Vegetation Index (NDVI) is a quick, effective indicator of plant life in a certain area. It is just a number that is calculated by comparing images taken in Red and Near-infrared (NIR) light. NDVI is most sensitive to changes in the amount of green biomass and chlorophyll content. Your imaging system is equipped with both a red and NIR camera, so you will be able to calculate the NDVI values and assess plant health for lands in your area.

  • What is NDVI?
  • Calculating NDVI
  • Cautions
  • Images

NDVI example

Where does it come from?

NDVI takes advantage of the plant adaptation that causes healthy plants to reflect red light and near-infrared (NIR) light in extremely different ways. This adaptation is a result of how plants use sunlight for photosynthesis. The breaking up of proteins that occurs duirng photosynthesis requires light with enough energy. Red light and more energetic light has enough energy to power photosynthesis, so it is highly absorbed as the plant thrives. Near infrared light is not energetic enough to create organic compounds, but does contain enough energy to heat the plant up, which could lead to overheating and plant death. Since the plants can’t use NIR light and it could be somewhat harmful to them, they reflect it back. This means that healthy plants absorb very different amounts of red light and NIR light. Unhealthy or dead plants, however, are not performing photosynthesis, so they do not absorb as much energetic light and reflect it back. This means that unhealthy or dead plants will not have a notable difference in the amount of red and NIR light they reflect. Healthy Plant Spectrum

This graph shows the drastic difference between Red and NIR light reflected by a healthy plant. The blue line shows how the percentage of reflected light varies with wavelength. First note the small bump on the left of the graph around 550nm. This corresponds to green light, which makes sense since most plants have a green color. At longer wavelengths around 650nm-750nm, the curve dips down. This corresponds to red light, indicating that the plant is reflecting very little red light, even less than 5%! The NIR part of the spectrum is between 750nm and 800nm, and you can see a sharp jump to high reflectance in this region, indicating the plant is highly reflective in NIR, almost 60% - a HUGE difference from the red reflectance.

For more information about NDVI, check out these websites: Earth Observatory, Wikipedia

How to calculate NDVI:

NDVI equationThe NDVI index is just a number calculated by the ratio of the difference in Red and NIR reflectance detected in the image compared to the sum. The RED and NIR values need to be normalized. Normalizing means that measurements will not be sensitive to how large of an area that is being imaged or how sensitive the cameras are to changes in height. To do this, we consider percentages, or decimals. This is where the black and white calibration panel comes in. Your red and NIR cameras do not measure the red and NIR reflectivity of a surface directly - it only measures Data Number (DN), which is the number of photons it receives. Your calibration panel helps you link DN to reflectance, so that when you measure a certain DN, you can read off the reflectance value as a percentage from your calibration graph. See the imaging page for details about how this works.

What the NDVI number actually means:

You can see from the equation, that the NDVI value will always fall between -1 and 1. For example, 0% NIR detection and 100% RED detection would give a value of -1. Negative values close to -1 (where NIR light is mostly absorbed and RED light is mostly reflected) usually correspond to water, like streams or lakes. Numbers that are close to zero, like -0.1 and +0.1 indicate that RED and NIR photons are being received in relatively equal amounts by the detector, which usually corresponds to barren ground, like rock, sand, or snow. Low positive numbers from 0.2 to 0.4 indicate that more NIR is being detected, corresponding to some plant life, like grass or shrubs. Very high positive values close to +1 indicate that almost all the light received is in the NIR and very little in the RED, corresponding to densely forested land, like rainforests.

You should be aware that NDVI is not a perfect measure of plant health; it can be sensitive to a variety of influences.

Cloud ShadowHigh altitude detectors, such as those in space, can look at a large area at once, but they must look through clouds and a lot of the atmosphere, which can significantly alter the data. If viewing a smaller patch of land, it’s much better to keep the detector lower, like with tethered balloon, so that clouds and atmosphere are less of a factor.

It’s important to take data on a clear day. Shadows from clouds can filter out the bright light that gives your images such good quality or even cause a different reading on the detector.

Soil effects - soil can darken when wet. If the NIR and RED ratios don’t change in the same way as dry soil, it can appear that there has been a change in vegetation when there really hasn’t.


Since the Sun is at different altitudes in the sky at different times of the day, it will reflect off the plant leaves in different ways. And the ground and surfaces of plants are not flat, causing some of the light will reflect off in different direction. This can easily add error to your data and reducing consistency and reliabiility. To minimize these effects, it is very important to take the NIR and RED data at the same time and from the same place. The parallax introduced by the two cameras being an inch or two apart can be corrected with your computer software, but you can help minimize error by flying at a consistent location and in a consistent orientation.

Different instruments can have subtle variations, causing one to be more sensitive to one kind of light than the other, so it’s important to compare results from the same instrument.

Red and NIR comparision photosHere you can see pictures taken in RED and NIR of branches with a background of thick, green leaves. The branches look very similar in RED and NIR; however, the background leaves are very different. The leaves are dark in RED, meaning they have absorbed a lot of the RED light and are reflecting back very little. In NIR, the leaves are bright - they are reflecting back a lot of NIR light and absorbing very little. We can use these images to estimate the NDVI value for this scene: the branches would probably have an NDVI close to zero, maybe on the negative side, like -0.1, since the branches are slightly darker in the NIR image. And the leaves would have a middle to high positive value, probably close to 0.5 or 0.6.

Global RED and NIR maps

Here are global images taken in RED (visible) and NIR filter bands. You can see that North America is very dark in RED and bright in NIR, so perhaps this photo was taken during the summer when many plants are in bloom and there is a lot of vegation. Look at the Saharan Desert in northern Africa - it is bright is both visible and infrared, which would give an NDVI close to zero. We know the desert is mostly sand, so this is exactly what we would expect.

What would you expect for areas near the equator that contain rainforests? What about near the arctic, which is made of ice? You can start to get a feel for how these two images can quickly and easily yield information about plant coverage and terrain.


North America low NDVINorth America high NDVIThese images plot the NDVI value over North America during different seasons, with white or tan representing a very low to negative NDVI and green representing a higher, positive NDVI. You can see how drastically the NDVI value can change from low to high vegetation cover.