Chap. 7 — Region Segmentation Region Segmentation

Region Segmentation

Computação Visual e Multimédia

10504: Mestrado em Engenharia Informática

Chap. 7 — Region Segmentation

Chapter 7: Region Segmentation

Outline

Chapter 7: Region Segmentation Image segmentation:��a reminder

Chapter 7: Region Segmentation Image segmentation:��… in pictures

Mea

n S

hift:

A R

obus

t App

roac

h to

war

d Fe

atur

e S

pace

Ana

lysi

s, b

y D

. Com

anic

iu a

nd P

. Mee

r ht

tp://

ww

w.c

aip.

rutg

ers.

edu/

~com

anic

i/MS

PAM

I/msP

amiR

esul

ts.h

tml

Edge segmentation:

Region segmentation:

http

://ro

bots

.sta

nfor

d.ed

u/cs

223b

/inde

x.ht

ml



What is a region?


Region-based approach


Region-based segmentation

€

Ri = Ii=1

n

€

Ri R j =∅ ∀i =1,2,…n

€

P Ri R j( ) = FALSE€

P(Ri) = TRUE, ∀i

Chapter 7: Region Segmentation Comparison of histogram, region growing ��and deformable contour segmentations

3 5 7 3 4 2 1 2 4 9 10 22 9 3 3 5 12 11 15 10 3 5 6 11 9 17 19 1 2 3 11 12 18 16 2 3 6 8 10 18 9 5 4 6 7 8 3 3 1

3 5 7 3 4 2 1 2 4 9 10 22 9 3 3 5 12 11 15 10 3 5 6 11 9 17 19 1 2 3 11 12 18 16 2 3 6 8 10 18 9 5 4 6 7 8 3 3 1

3 5 7 3 4 2 1 2 4 9 10 22 9 3 3 5 12 11 15 10 3 5 6 11 9 17 19 1 2 3 11 12 18 16 2 3 6 8 10 18 9 5 4 6 7 8 3 3 1

3 5 7 3 4 2 1 2 4 9 10 22 9 3 3 5 12 11 15 10 3 5 6 11 9 17 19 1 2 3 11 12 18 16 2 3 6 8 10 18 9 5 4 6 7 8 3 3 1

3 5 7 3 4 2 1 2 4 9 10 22 9 3 3 5 12 11 15 10 3 5 6 11 9 17 19 1 2 3 11 12 18 16 2 3 6 8 10 18 9 5 4 6 7 8 3 3 1

3 5 7 3 4 2 1 2 4 9 10 22 9 3 3 5 12 11 15 10 3 5 6 11 9 17 19 1 2 3 11 12 18 16 2 3 6 8 10 18 9 5 4 6 7 8 3 3 1

3 5 7 3 4 2 1 2 4 9 10 22 9 3 3 5 12 11 15 10 3 5 6 11 9 17 19 1 2 3 11 12 18 16 2 3 6 8 10 18 9 5 4 6 7 8 3 3 1

3 5 7 3 4 2 1 2 4 9 10 22 9 3 3 5 12 11 15 10 3 5 6 11 9 17 19 1 2 3 11 12 18 16 2 3 6 8 10 18 9 5 4 6 7 8 3 3 1

3 5 7 3 4 2 1 2 4 9 10 22 9 3 3 5 12 11 15 10 3 5 6 11 9 17 19 1 2 3 11 12 18 16 2 3 6 8 10 18 9 5 4 6 7 8 3 3 1

3 5 7 3 4 2 1 2 4 9 10 22 9 3 3 5 12 11 15 10 3 5 6 11 9 17 19 1 2 3 11 12 18 16 2 3 6 8 10 18 9 5 4 6 7 8 3 3 1

3 5 7 3 4 2 1 2 4 9 10 22 9 3 3 5 12 11 15 10 3 5 6 11 9 17 19 1 2 3 11 12 18 16 2 3 6 8 10 18 9 5 4 6 7 8 3 3 1

3 5 7 3 4 2 1 2 4 9 10 22 9 3 3 5 12 11 15 10 3 5 6 11 9 17 19 1 2 3 11 12 18 16 2 3 6 8 10 18 9 5 4 6 7 8 3 3 1

threshold T≥10 threshold T≥11 threshold T≥12

region growing with variance of 2 in respect to value 11 with reference to threshold T≥11

deformable contour to meet threshold T≥11


Region growing segmentation

seed growing final region

http

://ue

i.ens

ta.fr

/bai

llie

Chapter 7: Region Segmentation Seed-based region growing segmentation:��pixel aggregation

The seed point can be selected either by a human or automatically by avoiding areas of high contrast (large gradient) => seed-based method.

Chapter 7: Region Segmentation Seed-based region growing ��segmentation: example

original image

threshold: 225~255

threshold: 190~225

threshold = 255 returns multiple seeds

threshold: 155~255

http

://en

.wik

iped

ia.o

rg/w

iki/R

egio

n_gr

owin

g

Problem: To isolate the strongest lightning region of the image on the right hand side without splitting it apart. Solution: To choose the points having the highest gray-scale value which is 255 as the seed points shown in the image immediately below.

Chapter 7: Region Segmentation Fast scanning algorithm:��a kind of unseeded region segmentation

1

Threshold T: P1 == P2 iff Diff(Col(P1),Col(P2)) < T

val=? y

x x==y: val = x x<>y: boundary(x)y if |x-y|≤T new region index if |x-y|>T

1 2 2 3

1 1 1 1 1

If the criterion of homogeneity is local (compared to the value of the candidate pixel and the pixel of the border) => linear method.

http

://ue

i.ens

ta.fr

/bai

llie

Chapter 7: Region Segmentation Region growing segmentation:��advantages & disadvantages

http

://ue

i.ens

ta.fr

/bai

llie


Region splitting and merging segmentation

original image splitting & merging thresholding seg.

Chapter 7: Region Segmentation Region splitting: example��

original image split 1

split 2 split 3

In this example, the criterion of homogeneity is the variance of 1.

http

://ue

i.ens

ta.fr

/bai

llie


Splitting & merging: data structures

RAG with adjacency relations (in red) for big black region.


Splitting & merging segmentation algorithm

RAG with adjacency relations (in red) for big black region.

http

://as

tro.

tem

ple.

edu/

~si

ddu


Watershed segmentation

http

://en

.wik

iped

ia.o

rg/w

iki/W

ater

shed

_(im

age_

proc

essi

ng)#

cite

_not

e-1

watersheds

watershed crest line

Chapter 7: Region Segmentation Watershed segmentation ��by flooding

original image 3D topographic surface

This technique aims at identifying all the third type of points (i.e., points of watershed lines) for segmentation


Watershed segmentation by flooding

http

://eu

clid

.ii.m

etu.

edu.

tr/~

ion5

28/d

emo/

lect

ures

/6/4

/inde

x.ht

ml

255

0

255

0

255

0

255

0


Watershed segmentation algorithm

Chapter 7: Region Segmentation Watershed segmentation algorithm:��dam construction

Chapter 7: Region Segmentation Flooding-based watershed segmentation ��applied to gradient image

original image gradient image

watershed of the gradient image

final contours

http

://cm

m.e

nsm

p.fr

/~be

uche

r/w

tshe

d.ht

ml

Chapter 7: Region Segmentation Marker-controlled watershed ��segmentation (gradient image)

original image over-segmented image

http

://cm

m.e

nsm

p.fr

/~be

uche

r/w

tshe

d.ht

ml

markers of the blobs and of the background

marker-controlled watershed of the gradient image


Inter-pixel watershed segmentation

http

://en

.wik

iped

ia.o

rg/w

iki/W

ater

shed

_(im

age_

proc

essi

ng)#

cite

_not

e-2


More complex segmentation methods


Snakes

Copyright G.D. Hager Images taken from http://www.cs.bris.ac.uk/home/xie/content.htm


Level sets

Copyright G.D. Hager Images taken from http://www.cgl.uwaterloo.ca/~mmwasile/cs870/


Graph cuts

Copyright G.D. Hager Images taken from efficient graph-based segmentation paper

Chapter 7: Region Segmentation Generalized PCA��(Rene Vidal)

Copyright G.D. Hager

Human GPCA


Summary:

Documents

Chap. 7 — Region Segmentation Region Segmentation