Tutorials

Producing RGB images from co-registered data

This section will describe, step by step, the use of GSDView to create an RGB image. RGB stands for Red, Green and Blue. The technique consists in co-registering calibrated images (2 or 3) acquired at different times and storing them in the different channels of an RGB image (e.g. master in RED channel, slave1 in GREEN channel and slave2 in BLUE channel). Any changes in the scene result in coloured regions (caused by movement of objects, floods, land slides, etc.).

Open Image

To open a single image use the File menu or the File icon on the toolbar.

File menu

A window will open in which to select a file.

Open dialog

If the image has never been downloaded, you may have to wait until a raster band is available. The length of the wait will depend on the size of the image. Once downloaded you will be able to visualize the raster bands in the Dataset Browser menu. When a raster band is opened, an overview of the band and of the image itself will be shown.

Trasform into GTiff

Once the image is opened, the next step is to transform it into an 8 bit GTiff file. The Data Export tool, implemented by INNOVA, is ideal for this. It uses some GDAL functions, amongst which GDAL, gdal_translate and gdalinfo. Data Export can be configured by opening the Tools sub-menu, General.

General tools menu

The following parameters must be set in this window:

• Destination dataset: the name of the output file, including the path. By default the same name of the original file with an modified suffix and extension is given. If saving with desired name, be sure to use a file where you have writing permission.
• Output Data Type: the type of output data. Set to Byte

Output Format parameter is correctly set by default to GTiff.

Data export tool

Once the parameters are set accordingly, you can start the processing.

The GTiff file will be created and it can be visualized with any graphic interface. An example is given below:

Grayscale image (single channel)

Follow these two steps for each of the 3 co-registered images. You will obtain the same number of 8 bit GTiff images.

Please note that the Data Export tool, on top of carrying out the conversion of file format and data type, will also rescale the image to +/-3 standard deviations respect to the medium value of the data. This improves the readability of the data when using standard graphic interface software.

Merging

At this point you have 3 images of the same area. The GDAL gdal_merge.py tool comes in handy to merge them. To set the parameters, go to the Tools menu and click on the Misc To RGB Menu under the GDAL Tools Misc

Misc tools menu

The parameters must be set as follows:

• Source Dataset 1, 2, 3: represent the input data. Click on the checkbox for the first 3 data sets and insert the three 8 bit GTiff images generated beforehand.
• Output format:the format of the file that will be created. Set a GTiff
• Destination dataset: the name of the output file, comprehensive of path. By default the same name of the original file with an new suffix and extension is given. If saving with desired name, be sure to use a file where you have writing permission.
• Output Data Type: the type of output data. Set to Byte

Merge To RGB tool

At the end of the processing you can open the obtained image and check that it has 3 bands. The RGB image can be visualized with any graphic tool, showing the 3 original images, overlaid, each stored in a different colour band.

RGB image

You can Zoom in to see the changes between the images, shown in the various coloured regions. Following we show you the same area seen in each of the three original images.

Single channel image

And then seen in the final RGB image. You can clearly see the ships in transit.

RGB image (zoom)