Fak IV / Dept. ET-I  /  CG  /  Forschung  /  Fernerkundung  / 

Visualisierung von Fernerkundungsdaten

Ansprechpartner: Martin Lambers

Hintergrund

August 20, 2012: full source code for our Pacific Graphics short paper Ellipsoidal Cube Maps for Accurate Rendering of Planetary-Scale Terrain Data is available at http://ecm-planet.sourceforge.net/.

Visualization of remote sensing data involves many processing steps to convert sensor data to geometry and color. Moving as many of these computations as possible out of static preprocessing steps and into the interactively controlled visualization step allows explorative evaluation of multimodal remote sensing data.

In particular, an interactive visualization system for remote sensing data must fulfill the following requirements:

• It must allow interactive navigation in very large data sets.
• It must allow raw sensor data to be processed in an interactively adjustable way, so that the user can find optimal parameters for the task at hand.
• It must allow interactive combination of both geometry and texture data sets, to allow comparison and combined analysis of data sets from different sensors and/or different points in time.

In this research project, programmable graphics hardware is used to meet these requirements.
The two main areas of research currently are:

Visualization of Synthetic Aperture Radar (SAR) amplitude images

SAR images pose some unique challenges for interactive visualization:

• The dynamic range of the amplitude values is much larger than the dynamic range of common display devices. Specialized dynamic range reduction methods are required, both global and locally adaptive.
• Due to the coherent illumination used in SAR systems, SAR images are inherently affected by a high amount of multiplicative speckle noise. This may hinder the interpretation and analysis of the images. A visualization system must therefore offer adjustable despeckling methods.

Dynamic Terrain Rendering

Interactive processing and combination of Digital Elevation Models (DEMs) means that the geometry for the current view of the scene is not known a-priori. This prohibits the use of precomputed information to generate a continuous level of detail (CLOD). Instead, an adaptive mesh has to be built on the fly using only the dynamically generated height maps and little or no additional information.

Examples

	The well known NASA Blue Marble Next Generation (BMNG) data set. You are looking onto the University of Siegen.
	The BMNG data set, overlayed with a TerraSAR-X SAR image. The image shows the area around Rome, Italy.
	The same scene, with an activated lens that uses different global and local visualization parameters. The geometry is a mixture of the BMNG DEM and the CIA SRTM 4 DEM.
	A video of the same scene, zooming in on Rome and adjusting some visualization parameters. Note that interactive use of the application does not yet achieve the frame rate shown in the video.
	The same video with overlaid wireframe information, demonstrating the adaptivity of the mesh.
	The application running in our Virtual Reality Lab.

This research project is partially funded by grant KO-2960-3/1,2 from the German Research Foundation (DFG).

Veröffentlichungen

·	M. Lambers, A. Kolb
	Dynamic Terrain Rendering
	In 3D Research, 1(4), 2010, pages 1-8
	[bib]  [pdf]

·	M. Lambers, A. Kolb
	Visual Assistance Tools for Interactive Visualization of Remote Sensing Data
	In Proc. Int. IEEE Geoscience and Remote Sensing Symposium (IGARSS), 2010, pages 4745-4748
	[bib]  [pdf]

·	M. Lambers, A. Kolb
	GPU-Based Framework for Distributed Interactive 3D Visualization of Multimodal Remote Sensing Data
	In Proc. Int. IEEE Geoscience and Remote Sensing Symposium (IGARSS), 4, 2009, pages IV-57 - IV-60
	[bib]  [pdf]

·	M. Lambers, A. Kolb
	Automatic Point Target Detection For Interactive Visual Analysis of SAR Images
	In Proc. Int. IEEE Geoscience and Remote Sensing Symposium (IGARSS), 2, 2008, pages II-903 - II-906
	[bib]  [pdf]

·	M. Lambers, H. Nies, A. Kolb
	Interactive Dynamic Range Reduction for SAR Images
	In Geoscience and Remote Sensing Letters, 5(3), 2008, pages 507-511
	[bib]  [pdf]

·	M. Lambers, A. Kolb
	Adaptive Dynamic Range Reduction for SAR Images
	In Proc. 7th European Conference on Synthetic Aperture Radar (EUSAR), 3, 2008, pages 371-374
	[bib]  [pdf]

·	M. Kalkuhl, P. Droste, W. Wiechert, H. Nies, O. Loffeld, M. Lambers
	Parallel Computation of Synthetic SAR Raw Data
	In Proc. Int. Geoscience and Remote Sensing Symposium (IGARSS), 2007, pages 536-539
	[bib]

·	M. Lambers, A. Kolb, H. Nies, M. Kalkuhl
	GPU-based Framework for Interactive Visualization of SAR Data
	In Proc. Int. IEEE Geoscience and Remote Sensing Symposium (IGARSS), 2007, pages 4076 - 4079
	[bib]  [pdf]

·	Martin Lambers
	Interactive Visualization of Remote Sensing Data
	In PhD Thesis , University of Siegen, 2011
	[bib]  [pdf]

·	M. Lambers, A. Kolb
	Ellipsoidal Cube Maps for Accurate Rendering of Planetary-Scale Terrain Data
	In Proc. Pacific Graphics (Short Papers), 2012
	[bib]  [pdf]