2.2.32. Image classification

Performs automatic classification of an image (e.g. satellite image) based on a vector layer with training polygons. Random Forest algorithm is used.

You’ll need to input two files:

  • Image in GDAL-compatible format (preferably GeoTIFF). It can contain any number of bands, all of which will be used for classification.

  • Polygon vector layer with training objects. Any OGR-compatible vector format (preferably GeoPackage) is supported. Each polygon should have a field with object class number in its attributes.

Set up the parameters:

  • Name of the attribute that describes the class number for each object in the training polygons layer. The attribute must be an integer.

  • The number of trees in the decision forest. Leave blank to use the default value (100).

  • Maximum tree depth. Leave blank to not limit the tree depth.

Output:

  • Classified raster.

  • Classification report. It records the overall quality of the classification, the number of misclassified pixels in each class (and indicates the classes to which they were put instead).

Watch the video on youtube.

Launch instrument: https://toolbox.nextgis.com/operation/image_classification

Try it out using our sample:

Download input dataset to test the instrument. Step-by-step instructions included.

Get the output to additionally check the results.