Rendering Complexity in Computer-Generated Pen-and-Ink Illustrations

1. Rendering Complexity in Computer-Generated Pen-and-Ink Illustrations Brett Wilson & Kwan-Liu Ma The University of California, Davis

2. How would an artist treat this scene? • Ambiguous boundaries • Ambiguous depth

3. Method 1: Abstraction • Merge similar regions • Strokes don’t followgeometry exactly • Good color / texture

4. Method 2: Separation • Separate similar regions • Geometry is clear • Color is not as true

5. How can a computermake these decisions? • Introduction to NPR pipelines • Hybrid 2D / 3D pipeline • Abstraction: when and how? • Silhouette rendering • Hatching

6. Image-based NPR Image Image • Good abstraction • Low detail — always lose information

7. Geometry-based NPR Geometry Image • Poor abstraction • High detail — gain information

8. Geometry rendering techniques • Hierarchical textures [Winkenbach & Salesin 1994] • Arbitrary meshes [Girshick et al. 2000] • Smoothed direction fields [Hertzmann & Zorin 2000]

9. Neither of these techniquesworks well for complex scenes. • 2D approach gives too little detail, no relative importance • 3D approach gives too much detail, hard to pick out important things • Challenge: Intelligent use of abstraction

10. 2+D NPR processing • Hamel & Strothotte: Capturing and Re-Using Rendering Styles for NPR [EG ’99] • Generate multiple renderings • Match image attributes to example input • Discard geometry

11. Tree rendering. • Deussen & Strothotte: Computer-Generated Pen-and-Ink Illustrations of Trees [SIGGRAPH 2000] • Generate 2D depth renderings to extract important feature lines of the foliage. • Requires complex areas (leaves) to be tagged.

12. A generalized hybrid pipeline. • Add rendering and segmentation to the middle of the pipeline.

13. Silhouette rendering • Generate a complexity map • Indicates regions of high geometric complexity • Simplify areas likely to be confusing

14. A complexity map generatedfrom an edge rendering. Silhouette rendering Complexity map • Many other ways to measure complexity

15. Silhouette image should match a grayscale rendering. Edges Target Too light Too dark

16. Resulting edge rendering • Use Deussen’s technique to keep edges in order of importance • Add occluded edgesfor darkening

17. Grayscale rendering with hatching. • Artists don’t draw every object with separate strokes • Small, similar objects grouped and use the same strokes • Apply based on complexity

18. Segmentation for hatching • Use segmentation to identify groups of strokes. • Depth • Angle • Color • Texture • …etc.

19. Notes on segmentation. • Much easier than general image segmentation • No image understanding necessary • Simple segmentation is acceptable • Region growing

20. Segmentation-based hatching with important silhouette lines.

21. Primate Chest Isosurface • 3.5 M triangles • High detail • Ambiguous area

22. Silhouette edges

23. Fully hatched rendering

24. The rendering is separated into complex and non-complex regions. Simple Complex

25. Hybrid pen/paint rendering • Hatching fornon-complex areas • Solid black shadingfor complex areas • Preserves feel whilesimplifying rendering

26. Close-up comparison of hybrid rendering

27. Sharp boundaries • Blur operationaffects boundaries • “Knock out” large objects • Future work:Clustering in Z?

28. Conclusion • Abstraction • When • How • How will the viewer perceive the scene? • Incorporate segmentation in 3D pipeline • Clearer, more artistically believable pictures

29. Future work • Better models • Higher-quality hatching • More rendering styles in general • More possibilities with segmentation

30. Thank You • Funded by the U.S. National Science Foundation under • ACI 9983641 (PECASE award) • ACI 0325934 (ITR) • ACI 0222991