In that tutorial, we will see how to georeference images. By images we mean raster images as described in GIS Fundamentals tutorials (please, consult them for further details).

Raster (grid) images include: aerial pictures, scanned paper map, sattelite images, among others. A raster image being grid based, each pixel of an image as x (column), y (row) coordinates in pixels. At this stage, if we want incorporate such image in a GIS, we don’t have any geographic reference and we are not able to overlay it with other already georeferenced GIS layers on our region of interest. They don’t share the same coordinate system.

In our case we will georeference an satellite image showing the ARCAL RLA 5/051 different sudy sites. Our task wil consist in associating to (column, row) coordinates, geographic coordinates as shown below:

georeferencing

This process is called georeferencing. To collect reference geographic coordinates, you need to identify on the image you want to georeference, noticeable points easily identifiable in the field (intersection of roads, isolated tree, …) and take accurate GPS measurements. Such points are called Ground Control Points (GCP).

QGIS allows to georeference images very easily. To do so we will use in this tutorial the following files:

/dss_course_dataset/qgis/4-georeferencing/el_dorado_argentina.jpeg

Accurate GPS coordinates of Ground Control Points are:

GCP1 (lon/lat): -66.120186,-33.524861
GCP2 (lon/lat): -66.053540,-33.524706
GCP3 (lon/lat): -66.051052,-33.567278
GCP4 (lon/lat): -66.136184,-33.56497

See video tutorial either on your local copy or on YouTube QGIS-Georeferencing

Note that the output format of our georeferenced raster file is ‘tiff’. This format also known as ‘geotiff’ is a very common raster file format allowing to embed in a single file the raster image and associated georeferencing information.



Published

08 January 2001

Tags