Exercise 1 - Landsat ETM - Continued
|  | Arithmetic average image | |
Arithmetic
As we have seen, shallow lakes, swamps, and wet sand spits are barely identifiable in the true-colour image of the Niger Inland Delta.
They are, however, seen clearly in false-colour images using at least one infrared channel. But these images reflect a view of the Earth in an unusual way with regards to colours.
The combination of a false-colour image and the lifelike display of a true-colour image is a very useful method for producing images that closely resemble colour photographs.
Open the LEOWorks programme. If you have not downloaded the Niger Inland Delta images yet, do so now.
Choose File>Open. A dialogue box will pop up. Choose the folder Niger and select the first image Nigerdelta_Landsat_Band_2.tif. Open Nigerdelta_Landsat_Band_4.tif and Nigerdelta_Landsat_Band_7.tif, too.
Choose Image>Combine from...>Red Green Blue. A pop-up menu will open. Select image Nigerdelta_Landsat_Band_7.tif for Red, Nigerdelta_Landsat_Band_4.tif for Green and Nigerdelta_Landsat_Band_2.tif for Blue, and click OK.
This new image is a false-colour combination of three greyscale pictures. Improve the raw data the same way you did for the 'True-Colour Combination' exercise.
Open the image Nigerdelta_Landsat_Band_321.tif.
Choose Image>Arithmetic... and select Nigerdelta_Landsat_Band_321.tif as Image #1 and Combined Image as Image #2. Choose Average as Operation and activate All channels.
The new image combines the advantages of the two input images. It is an intersection of pixel values. Compare all three images and note the advantages of this newly calculated image by giving examples. Look attentively at vegetation cover and water surfaces.
NDVI
The Normalized Difference Vegetation Index (NDVI) is used to show the amount of vegetation cover in remote sensing images. Vegetation will give a strong reflection in the 0.7 - 0.9 micrometres area, whereas it will give a weak reflection in the 0.6 - 0.7 area. Because the spectral signature of vegetation is so characteristic, the distinction between bare ground and green vegetation is normally quite clear. The differences in reflection in the visible and near infrared ranges can, as mentioned already, be used to determine photosynthesis and plant growth.
Pigments in leaf cells absorb most of the red and reflect most of the near infrared radiation from the Sun.
The calculation for Landsat NDVI is: (channel4 - channel3) / (channel4 + channel3)
Take another look at the true-colour image of the Niger Inland Delta and observe the vegetation. It is quite difficult to classify the vegetation in this area. A NDVI image may be helpful.
| | | NDVI image of the Niger Inland Delta |
Open the LEOWorks programme. If you have not downloaded the Niger Inland Delta images yet, do so now.
Choose File>Open. A dialogue box will pop up. Choose the folder Niger and select the first image Nigerdelta_Landsat_Band_3.tif. Open Nigerdelta_Landsat_Band_4.tif, too.
Choose Image>Arithmetic... and select Nigerdelta_Landsat_Band_4.tif as Image #1 and Nigerdelta_Landsat_Band_3.tif as Image #2 and choose NDVI as Operation.
The calculated NDVI image can be improved by colouring it. In LEOWorks, choose Image/Convert to .../Image with LUT. Double-click on the grey scale at the bottom of the image. In the LUT Editor window click Open LUT and select ndvi.pal. Click Open. Close the LUT Editor window and save the image.
The bright areas indicate a high chlorophyll density, which in turn indicate healthy vegetation.
Alternatively, use the tool Compute NDVI.
Which areas appear in that bright colour?
To answer this question, compare the NDVI image with the true- and false-colour images of the Niger Inland Delta.
Find out what is special by comparing the true-colour and the NDVI images.
| 
