This article provides recommendations for the handling of data in LuciadFusion, based on the type of the data. Data formats and database data types each have their own specific characteristics and requirements. Those characteristics can have an effect on visualization detail and performance. In this article, you can find guidelines for dealing with particular format traits, so that you can get the best results from that format in LuciadFusion.

It helps you determine whether to:

  • Serve data on-the-fly

  • Before serving the data, let LuciadFusion Studio pre-process it automatically

  • Before serving the data, pre-process it manually as a LuciadFusion Coverage first.

Serve data on-the-fly?

LuciadFusion allows you to serve data on-the-fly: serve data from LuciadFusion Studio without pre-processing it in any way. The data is instantly available for serving. On-the-fly serving saves you time and disk space.

For certain types of data structures, though, serving on-the-fly is not possible, or it is not recommended.

See Table 1, “Data type recommendations” to find out which format types you can serve on-the-fly.

Automatic pre-processing in LuciadFusion Studio

In some cases, LuciadFusion automatically pre-processes data into more efficient formats, such as:

OGC 3D Tiles standard

Organizes 3D geospatial content, such as point clouds and 3D mesh data, into a tiled and multi-leveled hierarchical structure for efficient data exchange. For more information, see the OGC 3D Tiles documentation. The Tiling Engine and the 3D Tiles Processing Engine both convert data into OGC 3D Tiles.

Cubemap panorama data

Organizes panoramic images, grouped together in panoramas that represent a continuous scene, into a tiled and multi-leveled set of cubemaps for efficient data exchange. For more information, see the Panoramic images documentation.

These formats are automatically pre-processed in LuciadFusion Studio:

You can monitor the automatic pre-processing jobs from the LuciadFusion Studio user interface, and the Studio REST API.

Manual pre-processing

If you are working with data that:

  • Is unsuitable for serving on-the-fly

  • Does not get pre-processed automatically in LuciadFusion Studio

you can pre-process it as LuciadFusion Coverages yourself. LuciadFusion Coverage is a special file format that stores data as pre-computed raster tiles so that they can be served very efficiently from a WMTS or LTS service. WMS and WCS services significantly benefit from this format as well. The LuciadFusion process of computing and storing data as raster tiles is called fusing. The Tiling Engine converts data into LuciadFusion Coverages.

Data is typically better suited for pre-processing to LuciadFusion Coverages if:

  • It does not have overview levels. If your raster data consists of just one level containing all the detail, users will quickly lose sight of it when they zoom out of the map. The data set will show up as a red hatching pattern. If you pre-process it manually, you can define lower-detail overview levels yourself, so that the overview levels are visualized during zoom-out.

  • You need to serve the data to many clients concurrently.

  • Your data set is composed of hundreds or thousands of files. Those may be small or large, depending on your system. By tiling and multi-leveling such data, you can serve only the data relevant for the requested level-of-detail.

See Table 1, “Data type recommendations” for more type-specific recommendations.

Pre-processing data to a LuciadFusion Coverage takes time and disk space. On top of that, the process needs to be repeated at least partially if the data changes.

Two tools allow you to process data as a LuciadFusion coverage:

  • The fuser sample: run fusion.engine.bat/.sh from the command line to start the fusion.engine.Fuser sample with instructions. You can use the sample for the integration of coverage pre-processing into scripts.

  • The Data Connectivity Manager: You can use the LuciadFusion Data Connectivity Manager application to pre-process data into coverages. For more information, see the Data Connectivity Manager user’s guide.

You also have access to a point cloud pre-processor sample: samples.fusion.pointcloud.PointCloudPreprocessorTool or fusion.pointcloud.bat/.sh. If you are adding your point cloud data to LuciadFusion Studio, you won’t need it, though. LuciadFusion Studio processes several point cloud formats automatically.

Deciding how to serve data based on its type

To help you decide, the following sections present guidelines for serving your data on-the fly or pre-processing it into coverages first, based on the data type.

Table 1. Data type recommendations
Data type Serve on-the-fly, pre-process manually, or pre-processed automatically Notes

Binz

Pre-processed automatically

Binz data is automatically pre-processed to OGC 3D Tiles in LuciadFusion Studio. For more information about Binz data, see Binz data support.

CADRG

Serve on-the-fly or pre-process manually

CADRG data sets typically comprise many files. Pre-processing the data into LuciadFusion Coverages is recommended if that is the case.

CIB

Serve on-the-fly or pre-process manually

CIB data sets typically comprise many files. Pre-processing the data into LuciadFusion Coverages is recommended if that is the case.

Cubemap

Serve on-the-fly

Panoramic image data from other formats are converted to the Cubemap data format for serving. For more information, see Serving panoramic image data.

DEM

Serve on-the-fly or pre-processed automatically (LTS)

If you select the LTS service type, DEM data will be automatically pre-processed to a LuciadFusion coverage. For more information, see Serving elevation data from LuciadFusion Studio to LuciadRIA.

DHM

Serve on-the-fly or pre-processed automatically (LTS)

If you select the LTS service type, DHM data will be automatically pre-processed to a LuciadFusion coverage. For more information, see Serving elevation data from LuciadFusion Studio to LuciadRIA.

DMED/ DTED

Serve on-the-fly or pre-processed automatically (LTS).

You can usually serve DTED/DMED data on-the-fly, but take care to work with the dmed index file only. DTED elevation data comes as a series of .dt0, .dt1 and .dt2 files in a special directory structure. The LuciadFusion crawler picks up those files, but for efficient visualization it only uses the index file, called dmed.

If your data does not contain such an index file, you can easily create one using a text editor. Make sure that the file sits next to the dted directory, in this structure:

dted/
text/
dmed

The shortest acceptable dmed file simply specifies the global bounds of the data set as a single line of text. For example, if the data spans from 43 to 47 degrees north, and from 8 to 13 degrees east, the file could look like this:
N43N47E008E013

Pre-processing into a coverage may be the best approach if overview levels are missing from the DTED data set, for example, when .dt2 files are available but there are no .dt1 and .dt0 files.
If you select the LTS service type, DMED/DTED data will be automatically pre-processed to a LuciadFusion coverage. For more information, see Serving elevation data from LuciadFusion Studio to LuciadRIA.

E57

Pre-processed automatically

E57 data sets can contain both point cloud and panoramic image data.

ECDIS

Serve on-the-fly or pre-process manually

The nature of the ECDIS data determines the best approach. For more information, see Handling ECDIS data and its updates.

ECW

Serve on-the-fly or pre-process manually

The ECW format has a good compression ratio, and allows for the production of thumbnails without pre-processing. If the data set consists of hundreds or thousands of files, pre-processing it into coverages may still be recommended, depending on your system.

ETOPO

Serve on-the-fly or pre-processed automatically (LTS)

If you select the LTS service type, ETOPO data will be automatically pre-processed to a LuciadFusion coverage. For more information, see Serving elevation data from LuciadFusion Studio to LuciadRIA.

GeoTIFF

Serve on-the-fly, or pre-process manually. Elevation data pre-processed automatically (LTS)

The GeoTIFF format has many ways of internally representing the data. If serving GeoTIFF data on-the-fly does not result in good performance, the internal data structure is probably sub-optimal and we recommend pre-processing the data into coverages.

On-the-fly performance is often compromised for the GeoTIFF format if:

  • The data set consists of many GeoTIFF files (several hundreds or thousands)

  • The file has an internal structure that is sub-optimal for on-the-fly visualization. The data is sub-optimal if it has not been divided into tiles, or if it contains so-called stripes: 1-pixel-high, full-width tiles.

If you serve GeoTIFF data on-the-fly, and apply an automatic scale range to the GeoTIFF data on the map, it may happen that the data becomes invisible when you zoom out on the map. If that happens, the data has not been multi-leveled. If you pre-process the data up front, the pre-processing step adds more levels to the data, and the data remains visible when you zoom out on the map. To test whether your data remains visible without pre-processing, you can open all files on the map in the DCM application, for example.

If you select the LTS service type and the GeoTIFF files contain elevation data, the files will be automatically pre-processed to a LuciadFusion coverage.

GRIB

Serve on-the-fly

GRIB weather data often contains information that varies over time, pressure level, or altitude. For example, it may contain temperature maps for a series of altitude values. We typically refer to that type of data as multi-dimensional data. You can serve it on-the-fly. Clients of these protocols, such as a Lucy-based application, allow the users to interpret the dimensions. They can interactively inspect the data at distinct times or altitudes.

JPEG2000

Serve on-the-fly or pre-process manually

The JPEG2000 format has a good compression ratio, and allows for the production of thumbnails without pre-processing. If the data set consists of hundreds or thousands of files, pre-processing into coverages may still be recommended, depending on your system.

KML

Serve on-the-fly

You can use KML or KMZ data like any other format, and access its rich styling capabilities. However, as the format is heavily geared towards client-side viewers, some of its capabilities are not compatible with OGC web services. For more information, see KML data support and limitations.

LAS, LAZ

Pre-processed automatically

LAZ point clouds are automatically pre-processed to OGC 3D Tiles in LuciadFusion Studio. For more information, see Serving point cloud data from LuciadFusion Studio.

MrSID

Serve on-the-fly or pre-process manually

The MrSID format has a good compression ratio, and allows for the production of thumbnails without pre-processing. If the data set consists of hundreds or thousands of files, pre-processing it into coverages may still be recommended, depending on your system.

NetCDF

Serve on-the-fly

NetCDF weather data often contains information that varies over time, pressure level, or altitude. For example, it may contain temperature maps for a series of altitude values. We typically refer to that type of data as multi-dimensional data. You can serve it on-the-fly. Clients of these protocols, such as a Lucy-based application, allow the users to interpret the dimensions. They can interactively inspect the data at distinct times or altitudes.

NITF

Serve on-the-fly or pre-process manually

We recommend pre-processing the data if serving it on-the-fly proves too slow. In that case, the internal structure of the NITF data set is probably sub-optimal. The NITF format may contain JPEG2000, regular JPEG imagery, or even uncompressed data. The performance depends on the data contents.

OBJ

Pre-processed automatically

A 3D Tiles Processing Engine in LuciadFusion Studio automatically converts 3D mesh data in the OBJ format to OGC 3D Tiles.

OSGB 3D meshes

Serve on-the-fly

Although you can serve OSGB 3D reality mesh data as OGC 3D Tiles on-the-fly, some automatic pre-processing occurs in LuciadFusion Studio to generate the necessary metadata.

Pegasus

Pre-processed automatically

A Cubemap processing engine in LuciadFusion Studio automatically converts Pegasus panoramic image data to the Cubemap format. For more information, see Serving panoramic image data.