OGC GeoPackage Tiles Writer Parameters

Database Connection

The path to the destination GeoPackage file.

Note  This is automatically populated if you have specified the writer dataset.
  • When unchecked (default) and a database already exists, then tables and features are written to the existing database.
  • When checked, any existing database is deleted before writing begins. Checking this option also enables the Template File parameter.
Warning  After a database is deleted, it cannot be recovered – all data is lost.

If a database does not exist when writing begins, a new database is created.

Version

Sets the GeoPackage version (application_id and user_version fields).

The data format of the binary tile data stored in the tile table.

Tile Format Options

Description

Default Value

Auto (default)

When the Tile Format is Auto, the writer will use:

  • PNG for color tiles with transparency, or for integer coverage tiles
  • JPEG for opaque color tiles
  • TIFF for floating-point coverage tiles

 

JPEG or WEBP

Setting the Tile Format to JPEG or WEBP enables the JPEG/WEBP Compression Level parameter. This sets the desired compression level of the output images. This is expressed as a percentage of the original file size. For example, setting a compression level of 75 means that the output image will be approximately 75% smaller (one-quarter of the size) than an uncompressed image.

Value: The level of compression that is used, from 1-100.

The gpkg_webp extension will be registered in the GeoPackage.

Note  Repeatedly reading and saving files may decrease the quality for this format. In addition, FME Workbench may not know which format data comes from, and whether it used lossy compression. If you know that source data comes from files with lossy compression, you should also set the compression level to 0. Although this compression still does not guarantee the exact copy of the source file, the output will be very close to the original.

 

PNG

Setting the Tile Format to PNG, enables the PNG Compression Level parameter. This sets the PNG DEFLATE compression level.

Must be a value from 1-9.

6

PNG8

Setting the Tile Format to PNG8 enables the PNG8 Dither parameter, which determines if the PNG8 data will use Floyd-Steinberg dithering. Floyd-Steinberg dithering is an algorithm that reduces the color palette of an image while preserving as much image detail as possible.

  • Yes – PNG8 data will be written using Floyd-Steinberg dithering.
  • No – Dithering will not be used.

No

TIFF

If Tile Format is set to TIFF, then the gpkg_2d_gridded_coverage extension will be registered in the GeoPackage, and only single-band Real32 rasters will be written.

 

Parameter

Description

Default

Tiling Scheme

To enable the use of GeoPackage with tile servicing software, you can select a predefined tiling scheme:

  • GoogleCRS84Quad – Consists of a single 256x256 tile at its zoom level 0, in EPSG:4326 CRS, with extent in longitude and latitude in the range [-180,180]. Due to unused lines at zoom level 0, InspireCRS84Quad or PseudoTMS_GlobalGeodetic is recommended.
  • GoogleMapsCompatible – Consists of a single 256x256 tile at its zoom level 0, in EPSG:3857 CRS, with extent in easting and northing in the range [-20037508.34,20037508.34].
  • InspireCRS84Quad – Consists of two 256x256 tiles at its zoom level 0, in EPSG:4326 CRS, with extent in longitude in the range [-180,180] and in latitude in the range [-90,90].
  • PseudoTMS_GlobalGeodetic – Has the same definition as the InspireCRS84Quad tiling scheme. Note, however, that full interoperability with TMS is not possible due to the origin of numbering of tiles being the top left corner in GeoPackage, whereas TMS uses the bottom left corner as origin.
  • PseudoTMS_GlobalMercator – Consists of four 256x256 tiles at its zoom level 0, in EPSG:3857 CRS, with extent in easting and northing in the range [-20037508.34,20037508.34]. Note, however, that full interoperability with TMS is not possible due to the origin of numbering of tiles being the top left corner in GeoPackage, whereas TMS uses the bottom left corner as origin.
  • Use Incoming Feature – Instead of a standard tiling scheme, uses incoming raster extents and resolution as basis for tiles.

In all the above tiling schemes, consecutive zoom levels defer by a resolution of a factor of two.

GoogleMapsCompatible

Interpolation Type

Determines the resampling algorithm used:

  • Bilinear – Use the Bilinear interpolation algorithm to construct the tile data. Calculates a pixel's value by averaging the closest 2x2 neighborhood of pixel values. This method produces a higher-quality result than Nearest Neighbor.
  • Nearest-neighbor – Use the nearest-neighbor interpolation to construct the tile data. Nearest-neightbor interpolation uses the value of the nearest pixel to calculate the value of a given pixel without considering the values of pixels outside this nearest pixel range. Nearest-neighbor is the least resource-intensive, but also produces the lowest-resolution tiles.
  • Cubic – Use the cubic interpolation algorithm to construct the tile data. Calculates a pixel's value by using the closest 4x4 neighborhood of pixel values.
  • Cubic-spline – Use the Cubic Spline interpolation algorithm to construct the tile data. Because it calculates a pixel's value by using a spline, this results in a smaller error than linear interpolation, and the interpolation is smoother.
  • Lanczos – Use the Lanczos interpolation algorithm to construct the tile data. Use Lanczos interpolation when you want to enhance or maintain local contrast; it is useful for viewing detailed features or boundaries.
  • Mode – Use the Mode resampling method to construct the tile data. This is a custom resampling algorithm made by GDAL. This method will select the value which appears most often in all the sampled pixels and use that as the value for the given pixel.
  • Average – Use the average resampling method to construct the tile data. This is a custom resampling algorithm made by GDAL. This method will use the average of all non-NODATA contributing pixels to compute the value of a pixel.

Bilinear

Base Zoom Level Strategy

Sets the strategy used to determine the zoom level:

  • Auto – Round the zoom level to the closest integer value when computing the resolution to determine the zoom level.
  • Upper – Round the zoom level down when computing the resolution to determine the zoom level. For example, round 4.25 to 5. This will lead to oversampling.
  • Lower – Round the zoom level up when computing the resolution to determine the zoom level. For example, round 4.75 to 4. This will lead to undersampling.

Auto

Generate Zoom Levels

Determines whether to create additional zoom levels by resampling the original raster:

  • Yes – The writer generates zoom levels from the level below the Base Zoom Level, as determined by the Base Zoom Level Strategy parameter. It continues to generate less-resolute zoom levels until it hits the level specified in the Minimum Zoom Level parameter (described below), or until it is impossible to generate less-resolute zoom levels. The interpolation type used to resample the generated zoom levels is controlled by the Interpolation Type parameter.
  • No – Only the base (closest to native image resolution) zoom level will be written.

No

Minimum Zoom Level

Determines the least-resolute zoom level to generate when Generate Zoom Levels is set to Yes.

0, which is the lowest possible zoom level

Tiled Gridded Coverages parameters only apply to single-band rasters with an interpretation that is one of:

  • UInt16
  • Int16
  • Real32

These rasters will be written as a Tiled Gridded Coverage, according to the gpkg_2d_gridded_coverage extension.

Parameter

Description

Default

Precision

The smallest cell value that has meaning for the tiled gridded coverage.

1

Grid Cell Encoding

Describes how the value of each cell is related to the bounds of the cell. Standard values include:

  • grid-value-is-center (sample point is center of cell)
  • grid-value-is-area (sample area is the entire cell)
  • grid-value-is-corner (sample point is corner of cell)

grid-value-is-center

Unit of Measure

The unit for cell values in the tiled gridded coverage. The value for this parameter should be a symbol from The Unified Code for Units of Measure. For example, Cel is used for degrees Celsius in a temperature coverage.

If no value is specified, then the unit from the raster’s coordinate system will be assumed.

None

Field Name

The name that describes the data in the tiled gridded coverage. For example, a coverage of air temperature may use the field name Temperature.

Height

Quantity Definition

The description of the field of data in the tiled gridded coverage. This parameter expands upon the Field Name to add extra context or a link to a more in-depth definition. For example, the Marine Metadata registry uses the following definition for an Air Temperature field: Air temperature is the bulk temperature of the air, not the surface (skin) temperature.

Height

Spatial

Coordinate systems may be extracted from input feature data sources, may come predefined with FME, or may be user-defined. FME allows different output and input coordinate systems, and performs the required coordinate conversions when necessary.

How FME Processes Coordinate Systems in the Workspace

If a coordinate system is specified in both the source format and the workspace, the coordinate system in the workspace is used. The coordinate system specified in the source format is not used, and a warning is logged. If a source coordinate system is not specified in the workspace and the format or system does not store coordinate system information, then the coordinate system is not set for the features that are read.

If a destination coordinate system is set and the feature has been tagged with a coordinate system, then a coordinate system conversion is performed to put the feature into the destination system. This happens right before the feature enters into the writer.

If the destination coordinate system was not set, then the features are written out in their original coordinate system.

If a destination coordinate system is set, but the source coordinate system was not specified in the workspace or stored in the source format, then no conversion is performed. The features are simply tagged with the output system name before being written to the output dataset.

For systems that know their coordinate system, the Coordinate System field will display Read from Source and FME will read the coordinate system from the source dataset. For most other input sources, the field will display Unknown (which simply means that FME will use default values). In most cases, the default value is all you'll need to perform the translation.

You can always choose to override the defaults and choose a new coordinate system. Select More Coordinate Systems from the drop-down menu to open the Coordinate System Gallery.

Changing a Reprojection

To perform a reprojection, FME typically uses the CS-MAP reprojection engine, which includes definitions for thousands of coordinate systems, with a large variety of projections, datums, ellipsoids, and units. However, GIS applications have slightly different algorithms for reprojecting data between different coordinate systems. To ensure that the data FME writes matches exactly to your existing data, you can use the reprojection engine from a different application.

Note  To use this feature, you must have a supported GIS application installed on your system (for example, Esri ArcGIS®).

To change the reprojection engine, Select Workspace Parameters > Spatial > Reprojection Engine. In the example shown, you can select Esri (but the selection here depends on your installed applications):

For more information, see the links below:

Advanced

Determines where to store the temporary database while processing tiles.

  • On Disk (Default) – A temporary file will be created to process partial tiles. This mode is more reliable, but may be slower.
  • In Memory – A temporary database will be created entirely within memory. This mode is usually faster, but may be more susceptible to failure due to lack of system resources.

Determines the size in megabytes of the cache for the On Disk temporary database. If temporary database location is In Memory, then this parameter will be ignored.

The default cache size is 2.

Tip  Writing speed may be improved if cache size is set to a larger amount, such as 1x or 2x the available RAM on the system.