GrabCut GrabCut
Interactive Foreground Extraction Interactive Foreground Extraction
using Iterated Graph Cuts using Iterated Graph Cuts
Carsten Rother Carsten Rother
Vladimir Kolmogorov Vladimir Kolmogorov
Andrew Blake Andrew Blake
Microsoft Research Cambridge-UK Microsoft Research Cambridge-UK
GrabCut GrabCut
Interactive Foreground Extraction Interactive Foreground Extraction
using Iterated Graph Cuts using Iterated Graph Cuts
Carsten Rother Carsten Rother
Vladimir Kolmogorov Vladimir Kolmogorov
Andrew Blake Andrew Blake
Microsoft Research Cambridge-UK
Microsoft Research Cambridge-UK
Photomontage Photomontage Photomontage Photomontage
GrabCut – Interactive Foreground Extraction 1
video
Problem Problem Problem Problem
GrabCut – Interactive Foreground Extraction 2
Fast &
Accurate ?
What GrabCut does What GrabCut does What GrabCut does What GrabCut does
User Inpu t
Result
Magic Wand
(198?)
Intelligent Scissors
Mortensen and Barrett (1995)
GrabCut
Regions Boundary Regions & Boundary
GrabCut – Interactive Foreground Extraction 3
Framework Framework Framework Framework
Input: Image
Output: Segmentation
Parameters: Colour ,Coherence Energy:
Optimization:
GrabCut – Interactive Foreground Extraction 4
Graph Cuts Graph Cuts
Boykov and Jolly (2001) Boykov and Jolly (2001)
Graph Cuts Graph Cuts
Boykov and Jolly (2001) Boykov and Jolly (2001)
GrabCut – Interactive Foreground Extraction 5
Image Image
Min Cut Min Cut
Cut: separating source and sink; Energy: collection of edges Min Cut: Global minimal enegry in polynomial time
Foreground Foreground
(source) (source)
Background Background
(sink)
(sink)
Iterated Graph Cut Iterated Graph Cut Iterated Graph Cut Iterated Graph Cut
User Initialisation
K-means for learning colour distributions
Graph cuts to infer the segmentation
?
GrabCut – Interactive Foreground Extraction 6
1 2 3 4
Iterated Graph Cuts Iterated Graph Cuts Iterated Graph Cuts Iterated Graph Cuts
GrabCut – Interactive Foreground Extraction 7
Energy after each Iteration Result
Gua ran
teed to
con verg e
Colour Model Colour Model Colour Model Colour Model
Gaussian Mixture Model (typically 5-8 components)
Foregrou nd &
Backgrou nd
Backgrou nd
Foregrou nd
Backgrou
G
ndGrabCut – Interactive Foreground Extraction 8
R
G
Iterated R
graph cut
Coherence Model Coherence Model Coherence Model Coherence Model
An object is a coherent set of pixels:
GrabCut – Interactive Foreground Extraction 9
Blake et al. (2004): Learn jointly
Moderately straightforward Moderately straightforward examples
examples
Moderately straightforward Moderately straightforward examples
examples
… GrabCut completes automatically
GrabCut – Interactive Foreground Extraction 10
Difficult Examples Difficult Examples Difficult Examples Difficult Examples
Camouflage &
Low Contrast Fine structure No telepathy
Initial Rectangl e
Initial Result
GrabCut – Interactive Foreground Extraction 11
Evaluation – Labelled Database Evaluation – Labelled Database Evaluation – Labelled Database Evaluation – Labelled Database
Available online: http://research.microsoft.com/vision/cambridge/segmentation/
GrabCut – Interactive Foreground Extraction 12
Comparison Comparison Comparison Comparison
GrabCut Boykov 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 13
User Input
Result
Summary Summary Summary Summary
Magic Wand (198?)
Intelligent Scissors
Mortensen and
Barrett (1995)
GrabCut Rother et al.
(2004) Graph
Cuts
Boykov and
Jolly (2001)
LazySnappi ng
Li et al.
(2004)
GrabCut – Interactive Foreground Extraction 22