SITIS Archives - Topic Details
Program:  SBIR
Topic Num:  N102-116 (Navy)
Title:  Geospecific Displacement Maps for Real Time, Stereoscopic Training Simulation
Research & Technical Areas:  Air Platform, Information Systems, Sensors, Human Systems

Acquisition Program:  Joint Strike Fighter, ACAT I
  Objective:  Develop an innovative automated process that unlocks the advantages of real time displacement maps (or other tessellation technology) combined with geospecific two and three dimensional source imagery for implementing high complexity terrain surface regions in virtual environments for real time training simulators.
  Description:  State-of-the-art graphics chips are increasing in performance and beginning to incorporate displacement mapping technology. A displacement map consisting of texture imagery plus a vertical displacement for every point in the image can be created from aerial/satellite mono and stereoscopic imagery. Such a displacement map of a terrain area could contain embedded geospecific surface features such as trees, forest canopies, bushes, undergrowth, man made structures, etc with high resolution in both visual as well as three dimensional representations. Displacement mapping, edge detection analyses, and other new tessellation technologies have the potential to allow rapid incorporation of geospecific regions in a virtual world for real time training simulation. COTS software already exists which can automatically determine three dimensional displacements from mono as well as stereoscopic imagery. Thus, the process has potential to automatically incorporate highly cluttered terrain surfaces - without hand modeling the unique geometry of each terrain surface object. Performance issues include any impediments to a complete end-to-end solution, sensor simulation issues, patch insertion issues, occulted data hole issues, multiple source data issues, modeler time and skill issues, compatibility issues.

  PHASE I: Conceptualize and design an innovative process for rapid incorporation of a geospecific area into a training simulator starting with source imagery/data. Identify technology development, risk, and performance issues related to training simulation. Demonstrate the feasibility of the concept.
  PHASE II: Develop, construct and demonstrate the operation of a prototype including the process to make the virtual environment database as well as computing the simulation imagery of multiple new sensors at real time rates suitable for pilot training. Demonstrate compatibility with real time training simulation stereo visual systems.

  PHASE III: Transition this new capability to both military and commercial training simulators. PRIVATE SECTOR COMMERCIAL POTENTIAL/

  DUAL-USE APPLICATIONS: Real time high performance simulation such as commercial aviation flight trainers would benefit from successful outcome of this topic. The visual image generation industry would also benefit. For example 3D virtual representation of stereo-microscopy images will allow analyses of microscopic surfaces by a virtual interactive fly-through of even microscopic imagery (a bacteria could appear as big as a house).

  References:   1. Ephanov A. & Coleman C. (2006). Virtual Texture: A Large Area Raster Resource for the GPU. Interservice/Industry Training, Simulation, and Education Conference (I/ITSEC). to order go to http://www.simsysinc.com/IITSEC/ABS2006/SIM2006.htm#_Toc152918112 2. Glandville R. (2004). Texture Bombing. GPU Gems (Ch. 20). Addison-Wesley Reading, MA. 3. Hertzmann A., Jacobs C., Oliver N., Curless B. & Salesin D. (2001). Image Analogies. SIGGRAPH 2001 Proceedings, Los Angeles, CA. http://doi.acm.org/10.1145/383259.383295 4. Tsai, F. & Lin, H. -C. (2007). Polygon-based texture mapping for cyber city 3D building models. International Journal of Geographical Information Science, 21 (9), (pp. 965-981) 5. Nayak, Shailesh. ISPRS TC IV: Geo-databases and Digital Mapping - Trends and Challenges. ISPRS VOL 10, (pp.18-20). http://www.isprs.org/publications/highlights/highlights0605/13HL0605Society.pdf 6. Wei, L. (2005). Tile-Based Texture Mapping, GPU Gems 2, Addison-Wesley, Reading, MA. 7. Bunnell, M. (2005). Adaptive Tessellation of Subdivision Surfaces with Displacement Mapping, GPU Gems 2, Addison-Wesley, Reading, MA. http://http.developer.nvidia.com/GPUGems2/gpugems2_chapter07.html 8. Ali, S., Jieping Ye, Razdan, A. & Wonka, P. (2009). Compressed Facade Displacement Maps. Visualization and Computer Graphics, IEEE Transactions on, 15(2) (pp. 262-273). http://ieeexplore.ieee.org/xpl/freeabs_all.jsp?arnumber=4585376

Keywords:  virtual; software; simulation; imaging; visual; database

Questions and Answers:
Q: 1. What types of sensors are considered for the simualation imagery?
2. Is the use of both 2D and 3D (LIDAR) data formats required? If demonstration is done with one or the other, would that suffice?
3. Does the Navy prefer any particular COTS software to generate displacement maps?
A: 1. High magnification sensors that create imagery which can be dependent on the types of materials in the scene. The proposed solution should not be incompatible with whatever approach is being used for modeling sensor imagery.
2. No. It might suffice. "Suffice" depends on a lot of details as described in the topic.
3. The technical approach for creating and running real time rate imagery based on displacement maps or other tessellation techniques is up to the proposer.

Record: of