Examples

Land Use and Land Cover

• How To: Land-Use-Land-Cover Prediction for Slovenia
  • Before you start
    • Requirements
    • Overview
      • Part 1:
      • Part 2:
• Part 1
  • 1. Define the Area-of-Interest (AOI):
    • Get country boundary
    • Split to smaller tiles and choose a 5x5 area
    • Visualize the selection
  • 2. - 4. Fill EOPatches with data:
    • Define some needed custom EOTasks
    • Define the workflow tasks
      • Help! I prefer to calculate cloud masks of my own!
      • Reference map task
    • Define the workflow
    • Visualize the patches
    • Visualize the reference map
    • Plot the map of valid pixel counts
    • Spatial mean of NDVI
    • Temporal mean of NDVI
    • Plot the average cloud probability
• Part 2
  • 5. Prepare the training data
    • Define EOTasks
    • Run the EOWorkflow over all EOPatches
  • 6. Model construction and training
    • Set up and train the model
  • 7. Validation
    • Plot the standard and transposed Confusion Matrix
    • ROC curves and AUC metrics
    • Most important features
  • 8. Visualization of the results
    • Define EOTasks
    • Define Tasks and the Workflow
    • Run the prediction and export to GeoTIFF images
    • Visualise the prediction
    • Visual inspection of patches

Water Monitor

• Measuring the water level of a Theewaterskloof Dam in South Africa
  • Imports
  • Water level extraction EOWorkflow
    • Load the Polygon of nominal water extent and define a BBox
    • Step 1: Intialize (and implement workflow specific) EOTasks
      • Create an EOPatch and add all EO features (satellite imagery data)
      • Burn in the nominal water extent
      • The cloud mask is already provided by Sentinel Hub, we use it to calulcate the valid data mask.
      • Apply Water Detection
    • Step 2: Define the EOWorkflow
    • Step 3: Run the workflow
• Plot results

IO Examples

• Example showcasing IO tasks
  • Prerequisites
    • Sentinel Hub account
    • Credentials
  • Request different types of layers and data collections to an eopatch
    • General parameters
    • Tasks
      • Task for Sentinel-2 L1C data
      • Task for retrieving NDVI index
      • Task for Digital Elevation Model Data
      • Task for Sentinel-2 L2A and SCL data
      • Tasks for saving/outputting the eopatch
    • Run workflow
    • Plot results
      • Sentinel-2 L1C RGB bands
      • NDVI data
      • NDWI data
      • Sentinel-2 L2A RGB bands
      • Sentinel-2 Scene Classification Layer (from Sen2cor)
      • Mapzen Digital Elevation Model
  • Load the saved eopatch from disk

Data masks

• Example of generation of valid data mask using a predicate
  • Workflow to merge the cloud mask with the mask defining where reflectance data is available
  • Run workflow
  • Plot results

Visualizations

• EOPatch visualization
  • Basics
  • Plotting filters and configurations
  • Types of plots

Timelapse

• Creating timelapse animations
  • 1. Define tasks
    • 1.1 Download Task
    • 1.2 Filter Task
  • 2. Create and execute workflow
  • 3. Create animation