Skip to main content

7. Classify

After creating your assessment points and training data, you're ready classify your Imagery. The classification process uses your AI model to recognize different vegetation types based on your ground-truth points.

Overview​

The Classify Imagery module in TytonAI allows users to run AI-based classification of vegetation types over raster imagery using trained models. This process leverages both fine-tuned and base machine learning models to generate polygon classifications, which are visualized directly on the map and stored within the project.

Steps to Classify Imagery​

1. Open the Classify Tab​

Navigate to the Classify Imgaery tab to begin the classification process.

2. Select Classification Settings​

Within the Run Classification panel:

  • Classify Polygon: Choose the spatial extent (usually a Training Area or Study Area) over which classification will run.
  • Model: Select the pre-trained model to be used.
  • Epoch: Choose the epoch version of the model.

The selected model will define the vegetation classes available for classification (e.g., Ground, Shrub, Tree, Tussock, etc.).

3. Run Classification​

Once configuration is complete:

  • Click Classify to initiate the classification job.
  • The progress will be visible in the Job History.
  • Once completed, classification results are saved to the map interface as a new raster layer under Study Area > Classification Results.

4. Import Pre-existing Classification (Optional)​

You can also import a previously generated classification file:

In the Import Classification section:

  • Provide a Results Name.
  • Select the Study Area destination.
  • Choose one or more raster files containing classification data.
  • Optionally, skip adding to the project by enabling Skip Add to Project.

Click Import Classification to upload and visualize the layer.

5. Review Results​

Once classification is complete or imported:

  • The results are added as a new layer on the map.
  • Right-click on the result in the Study Area or Layer Panel to:
    • Rename.
    • Zoom to layer.
    • Apply filters or merge classifications.

Each class is visually distinguished by color overlays, enabling spatial vegetation pattern analysis.

Downloading the Classification as a TIF File​

Once your classification has been successfully generated and displayed in the project, you can export it for later use or to apply it in a different project:

  1. In the Explorer panel on the left, navigate to:
    • Study Area → Classification Results → [your classification result]
  2. Right-click the classification result layer.
  3. Click Download File(s) from the context menu.
  4. This will export the classification as a .tif raster file.

Use Cases for Exporting​

  • Backup: Store a local copy of the classification output.
  • Portability: Upload the .tif file to a different project without retraining the model.
  • Analysis: Import the .tif into external GIS software (such as QGIS or ArcGIS) for further analysis or visualization.

You can re-upload this .tif at any time using the Import Classification feature in the Classify Imagery module.

tip
  • Ensure that the selected training area and model are compatible in terms of spatial resolution and class coverage.
  • The classification layer can be used in further analysis like Accuracy Assessment and Vegetation Reporting.