• 沒有找到結果。

Graph Cut

N/A
N/A
Protected

Academic year: 2022

Share "Graph Cut"

Copied!
36
0
0

加載中.... (立即查看全文)

全文

(1)

Graph Cut

Digital Visual Effects Yung-Yu Chuang

with slides by Fredo Durand, Ramesh Raskar

(2)

Graph cut

(3)

Graph cut

• Interactive image segmentation using graph cut

• Binary label: foreground vs. background

• User labels some pixels

– similar to trimap, usually sparser

• Exploit

– Statistics of known Fg & Bg – Smoothness of label

• Turn into discrete graph optimization

– Graph cut (min cut / max flow)

F

B F

F F F B

B B

(4)

Energy function

• Labeling: one value per pixel, F or B

• Energy(labeling) = data + smoothness

– Very general situation – Will be minimized

• Data: for each pixel

– Probability that this color belongs to F (resp. B) – Similar in spirit to Bayesian matting

• Smoothness (aka regularization):

per neighboring pixel pair

– Penalty for having different label – Penalty is downweighted if the two

pixel colors are very different – Similar in spirit to bilateral filter

One labeling (ok, not best)

Data

Smoothness

(5)

Data term

• A.k.a regional term

(because integrated over full region)

• D(L)=

i -log h[Li](Ci)

• Where i is a pixel

Li is the label at i (F or B), Ci is the pixel value

h[Li] is the histogram of the observed Fg (resp Bg)

• Note the minus sign

(6)

Hard constraints

• The user has provided some labels

• The quick and dirty way to include

constraints into optimization is to replace the data term by a huge penalty if not respected.

• D(L_i)=0 if respected

• D(L_i)=K if not respected

– e.g. K=- #pixels

(7)

Smoothness term

• a.k.a boundary term, a.k.a. regularization

• S(L)=

{j, i} in N B(Ci,Cj) (Li-Lj)

• Where i,j are neighbors

– e.g. 8-neighborhood

(but I show 4 for simplicity)

• (Li-Lj) is 0 if Li=Lj, 1 otherwise

• B(Ci,Cj) is high when Ci and Cj are similar, low if there is a discontinuity between those two pixels

– e.g. exp(-||Ci-Cj||2/22) – where  can be a constant

or the local variance

• Note positive sign

(8)

Optimization

• E(L)=D(L)+ S(L)

•  is a black-magic constant

• Find the labeling that minimizes E

• In this case, how many possibilities?

– 29 (512)

– We can try them all!

– What about megapixel images?

(9)

Labeling as a graph problem

• Each pixel = node

• Add two nodes F & B

• Labeling: link each pixel to either F or B

F

B

Desired result

(10)

Data term

• Put one edge between each pixel and F & G

• Weight of edge = minus data term

– Don’t forget huge weight for hard constraints – Careful with sign

B

F

(11)

Smoothness term

• Add an edge between each neighbor pair

• Weight = smoothness term

B

F

(12)

Min cut

• Energy optimization equivalent to min cut

• Cut: remove edges to disconnect F from B

• Minimum: minimize sum of cut edge weight

B

F

cut

(13)

Min cut <=> labeling

• In order to be a cut:

– For each pixel, either the F or G edge has to be cut

• In order to be minimal

– Only one edge label per pixel can be cut (otherwise could

be added)

B

F

cut

(14)

Energy minimization via graph cuts

Labels (disparities)

d1 d2 d3

edge weight

edge weight

) , ,

(x y d3 D

) , (d1 d1 V

(15)

• Graph Cost

– Matching cost between images – Neighborhood matching term

– Goal: figure out which labels are connected to which pixels

d1 d2 d3

Energy minimization via graph cuts

(16)

Energy minimization via graph cuts

d1 d2 d3

• Graph Cut

– Delete enough edges so that

• each pixel is (transitively) connected to exactly one label node

– Cost of a cut: sum of deleted edge weights

– Finding min cost cut equivalent to finding global minimum of energy function

(17)

Computing a multiway cut

• With 2 labels: classical min-cut problem

– Solvable by standard flow algorithms

• polynomial time in theory, nearly linear in practice

– More than 2 terminals: NP-hard

[Dahlhaus et al., STOC ‘92]

• Efficient approximation algorithms exist

– Within a factor of 2 of optimal

– Computes local minimum in a strong sense

• even very large moves will not improve the energy – Yuri Boykov, Olga Veksler and Ramin Zabih,

Fast Approximate Energy Minimization via Graph Cuts, International Conference on Computer Vision, September 1999.

(18)

Move examples

Starting point

Red-blue swap move

Green expansion move

(19)

The swap move algorithm

1. Start with an arbitrary labeling

2. Cycle through every label pair (A,B) in some order

2.1 Find the lowest E labeling within a single AB-swap 2.2 Go there if E is lower than the current labeling

3. If E did not decrease in the cycle, we’re done Otherwise, go to step 2

Original graph

A B

AB subgraph

(run min-cut on this graph) B

A

(20)

The expansion move algorithm

1. Start with an arbitrary labeling

2. Cycle through every label A in some order

2.1 Find the lowest E labeling within a single A-expansion 2.2 Go there if it E is lower than the current labeling

3. If E did not decrease in the cycle, we’re done Otherwise, go to step 2

(21)

GrabCut GrabCut

Interactive Foreground Extraction using Iter Interactive Foreground Extraction using Iter

ated Graph Cuts ated Graph Cuts

Carsten Rother Carsten Rother

Vladimir Kolmogorov Vladimir Kolmogorov

Andrew Blake Andrew Blake

Microsoft Research Cambridge-UK

Microsoft Research Cambridge-UK

(22)

Demo

• video

(23)

Interactive Digital Photomontage

Aseem Agarwala, Mira Dontcheva, Maneesh Agrawala, Steven Drucker, Alex Colburn, Brian Curless, David Salesin, Michael Cohen, “Interactive Digital Photomontage”, SIGGRAPH 2004

• Combining multiple photos

• Find seams using graph cuts

• Combine gradients and integrate

(24)
(25)
(26)
(27)
(28)
(29)
(30)

actual photomontage

set of originals perceived

(31)

Source images Brush strokes Computed labeling

Composite

(32)

Brush strokes Computed labeling

(33)

Interactive Digital Photomontage

• Extended depth of field

(34)

Interactive Digital Photomontage

• Relighting

(35)

Interactive Digital Photomontage

(36)

Demo

• video

參考文獻

相關文件

 From a source vertex, systematically follow the edges of a graph to visit all reachable vertices of the graph.  Useful to discover the structure of

• vertex

Light rays start from pixels B(s, t) in the background image, interact with the foreground object and finally reach pixel C(x, y) in the recorded image plane. The goal of environment

• 下面介紹三種使用greedy algorithm產生minimum cost spanning

• 有向圖(directed graph)、無向圖(undirected

Graph Algorithms Euler Circuit Hamilton Circuit.. for Sprout 2014 by Chin

even if bound changes on non-binary variables are present, such a subset can be used like the conflict detecting clause in SAT to represent the conflict in the conflict graph..

Lance Williams: Brute Force in image space (shadow maps in 1978, but other similar ideas like Z buffer, bump mapping using textures and so on).. Completely