Raster Tiling

Clips rasters into tiles for machine learning algorithms.


Raster Tiling is a generic preprocessing block that clips rasters into rectangle-shaped tiles and makes them ready for consumption by analytics algorithms. Tiling is especially important for large raster datasets that should be broken up into more manageable pieces to improve performance. Output tiles are GeoTIFFs with the same bands and resolution as the input data. Simple data augmentation is also supported.

General InformationDescription
Block TypeProcessing (data preparation)
Supported Input TypesAOI Clipped, any georeferenced GeoTIFF
Resolutionidentical to the input
PerformanceSome ML algorithms can incur errors with tiles if they only support RGB.

Block Capabilities

This block supports the following input and output capabilities.


"input_capabilities": {
    "raster": {
      "up42_standard": {
        "format": "GTiff"


"output_capabilities": {
    "raster": {
      "up42_standard": {
        "format": "GTiff",
        "bands": ">",
        "resolution": ">",
        "sensor": ">",
        "dtype": ">",
        "processing_level": ">",
        "tile_width": "${tile_width}",
        "tile_height": "${tile_height}"
      "custom": {
        "match_extents": "${match_extents}"

Learn more about the documentation here.

End User License Agreement

View the End User License Agreement conditions in a new tab.