I have released a major update for the Semi-Automatic Classification Plugin for QGIS 2.0, version 2.3.0.
This is the changelog:
-Pre processing tab: clip multiple rasters at once with coordinates or shapefile
-Pre processing tab: Landsat (4, 5, 7, and 8) utility for automatic conversion
to TOA Reflectance and At-Satellite Brightness Temperature, and optionally DOS1
-Post processing tab: classification report in pixel, percentage and area
-added language checkbox for switching between English and locale (default
-Land cover change calculation now uses SAGA command changedetection
This version features a new Pre processing tab with utilities for preparing raster bands to classification.
In particular, I have added the ability to clip multiple raster at once, using coordinates or a shapefile.
|Clip multiple rasters tab|
Also, a new tab allows for the automatic conversion of Landsat bands (Landsat 4, 5, 7, and 8) to TOA Reflectance and At-Satellite Brightness Temperature (for thermal bands). Also, it allows for a simple atmospheric correction using the DOS1 method (Dark Object Subtraction 1), which is an image-based technique.
It requires the metadata file (MTL.txt) which is provided with Landsat images. This process automatically convert every .tif file that is found in a defined directory, if the file name ends with a band number. For more information about the conversion equations, see here.
A new Post processing tab allows for the creation of a report of land cover classification, providing the pixel count, the percentage and area of each class. This is also available as option during the classification process.
|Classification report tab|
Finally, the land cover change calculation now uses SAGA command changedetection, which is rapider than the previous python code.
The updated Semi-Automatic Classification Plugin is already available through the QGIS repository, or can be downloaded here.