Real-time Data Analysis And Rendering With HTML5 Canvas Using OpenLayers And GeoServer

Tom Kunicki (OpenGeo) with Dave Blodgett (U.S. Geological Survey)

09:00 on Saturday 21st September (in Session 52, starting at 9 a.m., Sir Clive Granger Building: A39)

Show in Timetable

Description: The emergence of HTML5 Canvas support in browsers creates an opportunity to shift rendering and analysis tasks from the serverĀ­side to the client web browser. Sample implementations will be presented that take advantage of an OpenLayers HTML5 canvas raster pipeline, using data vended by GeoServer using WPS and WMS, coupled with SLD rendering transforms and an integrated WMS tile cache.
Abstract:

The emergence of HTML5 Canvas support in browsers creates an opportunity to shift rendering and analysis tasks from the server­side to the client web browser. Traditional workflows of distributed geospatial applications implement filtering and styling on the server requiring frequent communication between the client and server with each modification to filters or symbology. By implementing high­throughput filtering and symbology rendering in the client web browser application responsiveness is significantly increased. Raster representations of features and coverages can be dynamically filtered and symbolized in the browser with functionality recently developed in OpenLayers. Recent developments in GeoServer enable customized rendering of feature attributes and coverage data into raster pixels by use of SLD utilizing WPS and Rendering Transformations. These rasters are then cacheable and in a format suitable for filtering and symbology rendering in the browser. With the newly developed HTML5 Canvas functionality in OpenLayers, these rasters can be dynamically filtered and symbolized in the browser. This technique is being applied at the USGS to provide real­time rendering of complex spatio­temporal hydrologic and climate model outputs with constant performance. Sample implementations will be presented that take advantage of an OpenLayers HTML5 canvas raster pipeline utilizing data vended by GeoServer using WPS and WMS coupled with SLD Rendering Transforms and an integrated WMS tile cache.