Remote Sensing
Remote Sensing
Remote sensing allows significant analysis. For this project, I have used Landsat 8 data type downloaded from the U.S. Geological Survey website. The data is open source, though downloading requires data ordering process. The project aims to analyse green areas and also, other land cover types in the part of Georgia by processing initial dataset. I am representing the main results of the analysis. Project is made entirely in ArcMap.
Landsat 8 dataset has 11 bands. This visual represents the appearance of initial dataset. The Landsat dataset is displayed as a greyscale image before processing.
Composition of the bands makes image classifications. With three visual primary colour bands (red, green, blue) is possible to create "True Colour" image, which is almost similar to the "real" image observed by a human. This map represents the example of the "True Colour" image.
This image is an example of "False Colour" composition with three bands (with different combinations): near-infrared (NIR), red and green, which means that vegetations have a high reflectance in the NIR. On this example of "False Colour" image, Vegetation and bare ground are distinguishable from each other - False colour composite image with different bands. Vegetation appears bright green while the bare ground is shown with reddish colours.
NDVI calculation
The abbreviation NDVI stands for the normalised difference vegetation index and identifies greenness areas on the land. The map shows the distribution of green places with high and low levels.
The map is an example of unsupervised image classification.
The map is an example of supervised image classification. The scatter plot shows the distribution of each class.