GDAL - Geospatial Data Abstraction Library

This type of spatial data source allows you to read a wide variety of raster data formats. GDAL (Geospatial Data Abstraction Library) is an open source library specializing in raster data conversion and processing.

There can be one or more source image files (mosaic of images). The images of the mosaic can overlap and there can be holes in the mosaic.

When the data source is updated, if the option is activated, a pyramid of lower resolution images is created by JMap Server and stored in the image cache. The image cache is then used for high performance production of images on the fly at specific resolutions. Depending on the initial image sizes and parameters, the image cache size can become voluminous. Typically, the total image cache is about the same size as the original images. Available disk space must be sufficient before updating this data source. It is possible that the data format selected will not require the creation of a pyramid. In this case, the parameter will be ignored.

Selected files	Select one or more image files to read. You have to select only the files containing the raster data (.tif, .jpg, etc.) and not the additional files (.xml, .prj, etc.). If many files are selected, they will be considered as a mosaic. The files must be georeferenced.
Cache path	The directory where JMap Server will store generated image files for the various resolutions. Default is JMAP_HOME/cache/raster.
Image interpolation	Method for producing images with reduced resolutions from images with higher resolutions. Default is Cubic and this method usually produces the best results.
Compute statistics (may take some time)	Statistics on image pixels simplify the analysis of these images. JMap extensions such as JMap Imagery use these statistics to work. Calculating statistics can take several minutes and this option should only be activated when necessary. If you do not need statistics, do not enable this option.
Create a pyramid	Activates the creation of an image pyramid with several resolutions. This greatly improves performance for producing images on the fly.
Pyramid image interpolation	Method used to produce pyramid images. The default value is Cubic and this method usually produces the best results.

Multiple Coordinate Reference Systems

JMap Server Lima introduces this new feature that allows to select multiple coordinate reference systems for a spatial data source (SDS) created from GDAL. This allows to prepare in advance several versions of the raster SDS in order to offer optimal performances when processing requests from different applications.

Once the SDS is created, the Reprojections section of its configuration interface allows to add (+), update () or delete () alternative projection systems to the original one. The table in the section shows the systems configured for the SDS.

When adding a projection system, the SDS is projected into the system and an image pyramid is created and stored in a single file.

When an application uses the SDS, the one projected with the appropriate projection system is used (the projection system must be specified in the parameters of the application's request to JMap Server). For example, JMap NG applications use the EPSG:3857 - WGS 84 / Pseudo-Mercator system, so the SDS projected in this system is used by JMap NG.

If the SDS is not projected in the appropriate system, it is reprojected on the fly which consumes JMap Server resources and compromises the application performance.

Each projection system added to the spatial data source creates a pyramid of images that is stored in a file. You must provide enough storage space for this data.