photomontage
DESCRIPTION
GrabCut Interactive Image (and Stereo) Segmentation Carsten Rother Vladimir Kolmogorov Andrew Blake Antonio Criminisi Geoffrey Cross [based on Siggraph 2004 and ECCV 2004]. GrabCut – Interactive Foreground Extraction 1. Photomontage. GrabCut – Interactive Foreground Extraction 2. - PowerPoint PPT PresentationTRANSCRIPT
![Page 1: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/1.jpg)
GrabCut GrabCut Interactive ImageInteractive Image
((andand Stereo) Stereo) SSegmentation egmentation
Carsten RotherCarsten RotherVladimir Kolmogorov Vladimir Kolmogorov
Andrew BlakeAndrew BlakeAntonio CriminisiAntonio Criminisi
Geoffrey CrossGeoffrey Cross [based on Siggraph 2004 and ECCV 2004][based on Siggraph 2004 and ECCV 2004]
![Page 2: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/2.jpg)
PhotomontagePhotomontage
GrabCut – Interactive Foreground Extraction 1
![Page 3: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/3.jpg)
Talk OutlineTalk Outline
Hard Image Segmentation: Fore- vs. BackgroundHard Image Segmentation: Fore- vs. Background
Soft Segmentation: Alpha Matting
Stereo Segmentation: Exploit Depth
GrabCut – Interactive Foreground Extraction 2
![Page 4: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/4.jpg)
Problem Problem
GrabCut – Interactive Foreground Extraction 3
Fast & Accurate ?
![Page 5: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/5.jpg)
What GrabCut doesWhat GrabCut does
User Input
Result
Magic Wand (198?)
Intelligent ScissorsMortensen and Barrett (1995)
GrabCut
Regions Boundary Regions & Boundary
GrabCut – Interactive Foreground Extraction 4
![Page 6: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/6.jpg)
FrameworkFramework
Input: Image
Output: Segmentation
Parameters: Colour ,Coherence
Energy:
Optimization:
GrabCut – Interactive Foreground Extraction 5
![Page 7: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/7.jpg)
Maximum a posteriori estimator (MAP):
• Gibbs Distribution of the MRF
same as
Energy – Probabilistic Energy – Probabilistic ViewView
GrabCut – Interactive Foreground Extraction 6
- log
![Page 8: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/8.jpg)
Graph CutsGraph Cuts - - Boykov and Jolly Boykov and Jolly (2001)(2001)
GrabCut – Interactive Foreground Extraction 7
ImageImage Min CutMin Cut
Cut: separating source and sink; Energy: collection of edges
Min Cut: Global minimal enegry in polynomial time
Foreground Foreground (source)(source)
BackgroundBackground(sink)(sink)
![Page 9: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/9.jpg)
Iterated Graph CutIterated Graph Cut
User Initialisation
K-means for learning
colour distributions
Graph cuts to infer the
segmentation
?
GrabCut – Interactive Foreground Extraction 8
![Page 10: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/10.jpg)
1 2 3 4
Iterated Graph CutsIterated Graph Cuts
GrabCut – Interactive Foreground Extraction 9
Energy after each IterationResult
Guaranteed to
converge
![Page 11: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/11.jpg)
Colour ModelColour Model
Gaussian Mixture Model Gaussian Mixture Model (typically 5-8 components)(typically 5-8 components)
Foreground &Background
Background
Foreground
BackgroundG
GrabCut – Interactive Foreground Extraction 10
R
G
RIterated graph cut
![Page 12: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/12.jpg)
Coherence ModelCoherence ModelAn object is a coherent set of pixels:
Error (%) over training set:
25
How do we choose ?
25
![Page 13: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/13.jpg)
Gaussian MRF:
Linear regression gives in closed-form
Pseudo-Likelihood:
Parameter Learning Parameter Learning (Blake (Blake 2004)2004)
GrabCut – Interactive Foreground Extraction 12
approximation
=
![Page 14: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/14.jpg)
A Gaussian MRF is not a realistic texture model
Gaussian? Gaussian!Real Image syntheticGMRF
Parameter Learning - Parameter Learning - ProblemsProblems
GrabCut – Interactive Foreground Extraction 13
![Page 15: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/15.jpg)
Moderately simple Moderately simple examplesexamples
… … GrabCut completes automaticallyGrabCut completes automatically GrabCut – Interactive Foreground Extraction 14
![Page 16: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/16.jpg)
Difficult ExamplesDifficult Examples
Camouflage & Camouflage & Low ContrastLow Contrast No telepathyNo telepathyFine structureFine structure
Initial Rectangle
InitialResult
GrabCut – Interactive Foreground Extraction 15
![Page 17: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/17.jpg)
Evaluation – Labelled Evaluation – Labelled DatabaseDatabase
Available online: http://research.microsoft.com/vision/cambridge/segmentation/
GrabCut – Interactive Foreground Extraction 16
![Page 18: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/18.jpg)
Comparison Comparison GrabCutBoykov and Jolly (2001)
Error Rate: 0.72%Error Rate: 1.87%Error Rate: 1.81%Error Rate: 1.32%Error Rate: 1.25%Error Rate: 0.72%
GrabCut – Interactive Foreground Extraction 17
User Input
Result
![Page 19: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/19.jpg)
ComparisonComparison
Trimap Boykov and Jolly
Error Rate: 1.36%
Input Image Ground Truth BimapGrabCut
Error Rate: 2.13%
GrabCut – Interactive Foreground Extraction 18
Error rate - modestly increase
User Interactions - considerable reduced
![Page 20: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/20.jpg)
Results Parameter Results Parameter LearningLearning
GrabCut – Interactive Foreground Extraction 19
![Page 21: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/21.jpg)
ComparisonComparison
Magic Wand (198?)
Intelligent Scissors Mortensen and Barrett (1995)
GrabCutRother et al. (2004)
Graph Cuts Boykov and Jolly (2001)
LazySnappingLi et al. (2004)
GrabCut – Interactive Foreground Extraction 20
![Page 22: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/22.jpg)
““Mixed pixels”: Combination of fore- and Mixed pixels”: Combination of fore- and background background
Alpha Mask: Proportion of fore- and background Alpha Mask: Proportion of fore- and background Natural Matting Problem: Determine alpha,F,B Natural Matting Problem: Determine alpha,F,B
from C from C
Under-determined System: 3 Equations and 7 unknowns
Digital MattingDigital Matting
GrabCut – Interactive Foreground Extraction 21
![Page 23: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/23.jpg)
1. Simple Alpha & Simple Colour 2. Difficult Alpha & Simple Colour
3. Simple Alpha & Difficult Colour
4. Difficult Alpha & Difficult Colour
Existing Methods
Human ?
GrabCut
GrabCut – Interactive Foreground Extraction 22
Transparency - Transparency - TaxonomieTaxonomie
![Page 24: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/24.jpg)
Border Matting Border Matting
Hard Segmentation Automatic Trimap Soft Segmentation
GrabCut – Interactive Foreground Extraction 23
to
![Page 25: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/25.jpg)
Input Bayes MattingChuang et. al. (2001)
Knockout 2Photoshop Plug-In
ComparisonComparison
GrabCut – Interactive Foreground Extraction 24
With no regularisation over alpha
Shum et. al. (2004): Coherence matting in “Pop-up light fields”
![Page 26: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/26.jpg)
Solve
Mean Colour Foreground
Mean ColourBackground
GrabCut – Interactive Foreground Extraction 25
Natural Image MattingNatural Image Matting
Ruzon and Tomasi (2000): Alpha estimation in natural images
![Page 27: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/27.jpg)
Noisy alpha-profile
Border MattingBorder Matting
GrabCut – Interactive Foreground Extraction 26
1
0
Foreground
Mix
Back-ground
Foreground Mix Background
Fit a smooth alpha-profile with parameters
![Page 28: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/28.jpg)
Result using DP Border Matting
DP
t
Dynamic ProgrammingDynamic Programming
GrabCut – Interactive Foreground Extraction 27
Noisy alpha-profile Regularisation
t+1
![Page 29: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/29.jpg)
GrabCut BorderGrabCut Border Matting -Matting - ColourColour
Compute MAP of p(F|C,alpha) (marginalize over Compute MAP of p(F|C,alpha) (marginalize over B)B)
To avoid colour bleeding use colour stealing To avoid colour bleeding use colour stealing (“exemplar based inpainting” – Patches do not work)(“exemplar based inpainting” – Patches do not work)
[Chuang et al. ‘01] Grabcut Border Matting
GrabCut – Interactive Foreground Extraction 28
![Page 30: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/30.jpg)
ResultsResults
GrabCut – Interactive Foreground Extraction 29
![Page 31: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/31.jpg)
Stereo Video + Stereo Video + SegmentationSegmentation
Criminisi et. al. (2003): 4-Plane DP to handle occlusions properly
Left Sequence Right Sequence
Disparity Sequence
GrabCut – Interactive Foreground Extraction 30
![Page 32: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/32.jpg)
Occusion, left and rightOccusion, left and right
GrabCut – Interactive Foreground Extraction 31
![Page 33: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/33.jpg)
Background SubstitutionBackground Substitution
Criminisi et. al. (2004): Remove boundary artefacts (SPS algorithm)
GrabCut – Interactive Foreground Extraction 32
![Page 34: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/34.jpg)
Object InsertionObject Insertion
GrabCut – Interactive Foreground Extraction 33
![Page 35: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/35.jpg)
Focus on ForegroundFocus on Foreground
GrabCut – Interactive Foreground Extraction 34
![Page 36: Photomontage](https://reader038.vdocuments.site/reader038/viewer/2022110215/56815911550346895dc63cdf/html5/thumbnails/36.jpg)
Conclusions & FutConclusions & Future ure WorkWork
GrabCut – powerful interactive extraction tool
Iterated Graph Cut based on colour and contrast
Regularized alpha matting by Dynamic Programming
Stereo and Segmentation give supportive information
How to solve the difficult hair problem ?
GrabCut – Interactive Foreground Extraction 35
[Argawall et.al.2004]