Poisson composi-ng 3

Transcription

1 1

2 naïve composi-ng 2

3 Poisson composi-ng 3

4 our result 4

5 Objec+ves Seamless composi+ng Robust to inaccurate selec+on Output Quality - Limit color bleeding Time- Performance - Efficient method 5

6 Related Work: Poisson Composi+ng Pas+ng gradients instead of pixels Pérez 03, Georgiev 06, Pros: Blends seamlessly Linear computa+on Cons: Bleeding visible Foreground to background bleeding Poisson Composi+ng Result 6

7 Related Work: L 1 Norm Introduced in shape from shading Reddy 09 Pros: Reduced bleeding Cons: Nonlinear - Computa*onally intensive L 1 Norm Result 7

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 composi+on Changing boundaries 8

9 Contribu+ons Conceal bleeding in textured areas Be_er gradient field at boundary Efficient linear scheme 9

10 Overview Problem Descrip-on Hiding Residuals with Visual Masking Genera-ng a low- curl boundary Results and Comparisons 10

11 Foreground Algorithm Overview Selec-on Background

12 Foreground Algorithm Overview Selec-on grad Background grad

13 Foreground Algorithm Overview Selec-on grad paste Target gradient field Background grad

14 Foreground Algorithm Overview Selec-on grad paste Target gradient field Background integrate grad Result 14

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

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

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 17

18 Overview Problem Statement Hiding Residuals with Visual Masking Genera-ng a low- curl boundary Results and Comparisons 18

19 Our Strategy Use the weights to locate integra+on residuals where they are less visible Exploit perceptual effect: visual masking human percep+on affected by texture. 19

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

21 Visual Masking Demo Texture hides low- frequency content 21

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

23 Design of the Weights smooth region: needs high weight to prevent bleeding textured regions: low weight is ok because bleeding less visible weights (white is high) 23

24 Es+ma+ng the Amount of Texture RGB gradient field 24

25 Es+ma+ng the Amount of Texture Small Gaussian convolu-on RGB gradient field Noise controlling func-on Large Gaussian convolu-on 25

26 Es+ma+ng the Amount of Texture small Gaussian large Gaussian gradient field g 26

27 Es+ma+ng the Amount of Texture Texture Map Output (white indicates more texture) 27

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

29 Hiding Residuals with Visual Masking Bleeding only in textured areas Poisson composi+ng With texture weights 29

30 Hiding Residuals with Visual Masking Residuals Poisson composi+ng With texture weights 30

31 Hiding Residuals with Visual Masking Reduced bleeding but not fully Poisson composi+ng With texture weights 31

32 Overview Problem Statement Hiding Residuals with Visual Masking Genera-ng a low- curl boundary Results and Comparisons 32

33 Boundary Problems Target field v integrable iff curl(v)=0 Only boundary has non- zero curl Consequence of target field construc+on (see paper) curl using standard strategy Our strategy: reducing the curl 33

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

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

36 Overview Problem Statement Hiding Residuals with Visual Masking Genera-ng a low- curl boundary Results and Comparisons 36

37 Results and Comparisons Copy- and- paste 37

38 Results and Comparisons Poisson Reconstruc+on : Pérez 03, Georgiev 06, 38

39 Results and Comparisons Maximum Gradient : Pérez 03 39

40 Results and Comparisons Diffusion : Agrawal 06 40

41 Results and Comparisons Lalonde 07 41

42 Results and Comparisons L 1 Norm : Reddy 09 42

43 Results and Comparisons Our Result 43

44 Results and Comparisons Our Result L_1 Norm Lalonde Diffusion Max Poisson Running Time (Seconds) 44

45 Results and Comparisons Poisson 45

46 Results and Comparisons Our Result 46

47 Results and Comparisons Poisson 47

48 Results and Comparisons Our Result 48

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

50 Conclusion Robust composi+ng using visual masking Be_er gradient field at boundary Efficient linear scheme Poisson Reconstruc+on Our Result 50

51 Special Thanks Todor Georgiev Kavita Bala George Dre_akis Aseem Agarwala Bill Freeman Tim Cho Biliana Kaneva Medhat H. Ibrahim Ravi Ramamoorthi This material is based upon work supported by the Na+onal Science Founda+on under Grant No and No

52 Conclusion Robust composi+ng using visual masking Be_er gradient field at boundary Efficient linear scheme Poisson Reconstruc+on Our Result 52

53 Results and Comparisons 53

54 Results and Comparisons Poisson 54

55 Results and Comparisons Our Result 55