Presentation is loading. Please wait.

Presentation is loading. Please wait.

Terrain Synthesis by Digital Elevation Models Howard Zhou, Jie Sun, Greg Turk, and James M. Rehg 2006.10.05.

Similar presentations


Presentation on theme: "Terrain Synthesis by Digital Elevation Models Howard Zhou, Jie Sun, Greg Turk, and James M. Rehg 2006.10.05."— Presentation transcript:

1 Terrain Synthesis by Digital Elevation Models Howard Zhou, Jie Sun, Greg Turk, and James M. Rehg 2006.10.05

2 Table of Contents 1.Introduction 2.Feature extraction 3.Feature matching and alignment 4.Patch stitching 5.Conclusion Introduction Feature extraction Feature matching Patch stitching Conclusion

3 Why? Numerous applications Landscape design Flight simulators Feature film special effects Computer games Introduction Feature extraction Feature matching Patch stitching Conclusion

4 Terrain synthesis Fractal model fBm - fractional Brownian motion (Mandelbrot 1982) Midpoint displacement, recursive subdivision …(Fournier 1982, Miller 1986, Voss 1985, Lewis 1987, Szeliski, et al. 1989) Erosion model Physical erosion simulation (Kelley, et al. 1988) Combination of Both Fractal terrains with erosion features (Musgrave et al. 1989) Most commercial landscaping software such as Terragen, Bryce, Vue d’seprit, and Mojoworld, etc. Previous Work Introduction Feature extraction Feature matching Patch stitching Conclusion

5 Limitation of previous terrain synthesis approaches Limited control by user (parameter tuning) Hard to capture real terrain style Previous Work Introduction Feature extraction Feature matching Patch stitching Conclusion

6 What If ? Introduction Feature extraction Feature matching Patch stitching Conclusion

7

8 Patch based texture synthesis Image quilting (Efros and Freeman 2001), Graphcut (Kwatra et al. 2003) Feature guided texture synthesis Image analogy (Hertzmann et al. 2000), Feature matching and deformation (Zhang et al. 2003, Wu and Yu 2004) Related Work Introduction Feature extraction Feature matching Patch stitching Conclusion

9 Terrain synthesis is not simply texture synthesis on height fields. Terrain synthesis must preserve global features such as ridges and valleys. Terrain synthesis must be globally controllable. Unlike general textures, terrain doesn’t have natural boundaries. Terrain synthesis is not texture synthesis Introduction Feature extraction Feature matching Patch stitching Conclusion

10 First example-based terrain synthesis User control via feature sketches. Feature-based approach to matching and placement of large curvilinear terrain features. Tree-ordered patch placement algorithm. Multiple terrain style. Our Contribution Introduction Feature extraction Feature matching Patch stitching Conclusion

11 Feature Extraction Extract important terrain features (valleys, ridges, …) Feature matching and deformation Match terrain features between user sketch and terrain data to find candidate patch Use deformation to align features Patch stitching Use graph cuts and Poisson interpolation to remove visible seams between neighboring patches Procedure Introduction Feature extraction Feature matching Patch stitching Conclusion

12 Flowchart Feature extraction Matching and deformation Patch stitching Introduction Feature extraction Feature matching Patch stitching Conclusion

13 Finding ridges and valleys Branches and Ends Path Features Chang’s PPA algorithm (Profile recognition and polygon breaking) Branch End Path Feature Extraction Introduction Feature extraction Feature matching Patch stitching Conclusion

14 Why PPA? Grand Canyon (shaded relief) Edge detection result PPA result Introduction Feature extraction Feature matching Patch stitching Conclusion

15 Target Connection Profile Recognition Polygon Breaking Branch Reduction PPA explained Introduction Feature extraction Feature matching Patch stitching Conclusion

16 Input In action

17 Profile Recognition In action

18 Polygon building In action

19 Polygon Breaking In action

20 Branch Reduction In action

21 Result In action

22 Feature placement (tree traversal) Introduction Feature extraction Feature matching Patch stitching Conclusion

23 Why is order important? Raster-scan patch placement (ncc) Tree traversal Introduction Feature extraction Feature matching Patch stitching Conclusion

24 Most of the time, the feature patches need alignment before they can be used. Thin plate spline mapping for feature deformation Two sets of corresponding feature points from feature matching Small deformation in terrain does not alter terrain style BranchEndPath Feature alignment Introduction Feature extraction Feature matching Patch stitching Conclusion

25 d : Deformation energy from TPS warping g : Graphcut seam cost f : Feature dissimilarity i : Other user specified constraints Feature Patch Matching Introduction Feature extraction Feature matching Patch stitching Conclusion

26 SSD- based search (accelerated) Fill the synthesized height map Non-feature placement Introduction Feature extraction Feature matching Patch stitching Conclusion

27 Graphcut Graphcut Textures: Image and Video Synthesis Using Graph Cuts (Kwatra et al. 2003) Introduction Feature extraction Feature matching Patch stitching Conclusion

28 Poisson interpolation Poisson image editing (Perez et al. 2003) Modify the gradient and reconstruct Introduction Feature extraction Feature matching Patch stitching Conclusion

29 Mount Jackson

30

31 Grand Canyon

32

33 Flathead National Forest, MT

34

35 Mount Vernon, KY

36

37

38 Middle Earth

39

40

41

42

43

44 Conclusion We’ve presented an image-based algorithm for terrain synthesis It provides user control by intuitive sketch It preserves terrain style embedded in the original height field Introduction Feature extraction Feature matching Patch stitching Conclusion

45 Cape Girardeau, MO (failed)

46

47 Results Show video

48 PPA in action Introduction Feature extraction Feature matching Patch stitching Conclusion


Download ppt "Terrain Synthesis by Digital Elevation Models Howard Zhou, Jie Sun, Greg Turk, and James M. Rehg 2006.10.05."

Similar presentations


Ads by Google