naïve compositing

2. naïve compositing

3. Poisson compositing

4. our result

5. Objectives Seamless compositing • Robust to inaccurate selection • Output Quality • - Limit color bleeding • Time-Performance • - Efficient method

6. Related Work: Poisson Compositing Pasting gradients instead of pixels Pérez 03, Georgiev06, … • Pros: • Blends seamlessly • Linear computation • Cons: • Bleeding visible • Foreground to background • bleeding Poisson Compositing Result

7. Related Work: L1 Norm Introduced in shape from shading Reddy 09 • Pros: • Reduced bleeding • Cons: • Nonlinear • - Computationally intensive L1 NormResult

8. Related Work: Moving Boundaries Move the boundaries to avoid bleeding Jia 06 • Pros: • Avoids bleeding • Cons: • We don’t want boundaries • to change • Changed composition Changing boundaries

9. Contributions • Conceal bleeding in textured areas • Better gradient field at boundary • Efficient linear scheme

10. Overview • Problem Description • Hiding Residuals with Visual Masking • Generating a low-curl boundary • Results and Comparisons

11. Algorithm Overview Foreground Selection Background

12. Algorithm Overview Foreground Selection grad Background grad

13. Algorithm Overview Foreground Targetgradient field Selection grad paste Background grad

14. Algorithm Overview Foreground Targetgradient field Selection grad paste Background integrate grad Result

15. Standard Approach: Poisson compositing We seek image I with gradients close to target v. target gradient field =

16. Standard Approach: Poisson compositing We seek image I with gradients close to target v. target gradient field = Least-squares: output gradient field

17. Our Approach We seek image I with gradients close to target v. target gradient field = value that minimizes curl Weighted least-square: weight output gradient field

18. Overview • Problem Statement • Hiding Residuals with Visual Masking • Generating a low-curl boundary • Results and Comparisons

19. Our Strategy • Use the weights to locate integration residuals where they are less visible • Exploit perceptual effect: visual masking • human perception affected by texture.

20. Visual Masking Demo Can you spot all the dots?

21. Visual Masking Demo Texture hides low-frequency content

22. Visual Masking Demo Can you spot all the dots?

23. Design of the Weights smooth region: needs high weightto prevent bleeding textured regions: low weight is okbecause bleedingless visible weights (white is high)

24. Estimating the Amount of Texture RGB gradient field

25. Estimating the Amount of Texture Small Gaussian convolution Noise controlling function RGB gradient field Large Gaussian convolution

26. Estimating the Amount of Texture small Gaussian large Gaussian gradient field g

27. Estimating the Amount of Texture Texture Map Output (white indicates more texture)

28. Weight Formula = 1 – • v depends only on foreground and background • does not depend on the unknown I • weights are constant in the optimization • our energy is a classical least-squares optimization Weight output

29. Hiding Residuals with Visual Masking Bleeding only in textured areas With texture weights Poisson compositing

30. Hiding Residuals with Visual Masking Residuals With texture weights Poisson compositing

31. Hiding Residuals with Visual Masking Reduced bleeding but not fully With texture weights Poisson compositing

32. Overview • Problem Statement • Hiding Residuals with Visual Masking • Generating a low-curl boundary • Results and Comparisons

33. Boundary Problems • Target field vintegrableiffcurl(v)=0 • Only boundary has non-zero curl • Consequence oftarget field construction(see paper) • Our strategy: reducing the curl curl usingstandard strategy

34. Naïve Approach • Minimize • Unfortunately, the result is not seamless: output of naïve approach close-up

35. Our Approach: Least Squares Trade-off • Unknowns: target field v along boundary • Weights depend on texture (detail in paper) zero curl weight seamless compositing

36. Overview • Problem Statement • Hiding Residuals with Visual Masking • Generating a low-curl boundary • Results and Comparisons

37. Results and Comparisons Copy-and-paste

38. Results and Comparisons Poisson Reconstruction : Pérez 03, Georgiev 06, …

39. Results and Comparisons Maximum Gradient : Pérez 03

40. Results and Comparisons Diffusion : Agrawal 06

41. Results and Comparisons Lalonde 07

42. Results and Comparisons L1 Norm : Reddy 09

43. Results and Comparisons Our Result

44. Results and Comparisons

45. Results and Comparisons Poisson

46. Results and Comparisons Our Result

47. Results and Comparisons Poisson

48. Results and Comparisons Our Result

49. Discussion • Several parameters • - all examples use the same parameters • Discoloration may happen • - happens in all gradient based operators • Our model of visual masking is simple • - good for performance • - more complex model could be used

50. Conclusion • Robust compositing using visual masking • Better gradient field at boundary • Efficient linear scheme Poisson Reconstruction Our Result