dip chapter 3- mathematical morphology- binary

8
1/21/2009 1 C ha pt er 3 : Mathe m a t i ca l morphology Pa rt 1 : bi na r y m or ph ol og y Dr. Hojeij youssef Di g i ta l i m a g e proce s si ng 1 Bi na r yima g e Ba si c Dilation Erosion Open Close Ne i g hbortr a ns form  T h in an d t h ic k Geode si cre constr ucti on Mor phology i n g ray level Outline Dr. Hojeij youssef 2

Upload: firas-abdel-nour

Post on 14-Apr-2018

218 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: DIP Chapter 3- Mathematical Morphology- Binary

7/29/2019 DIP Chapter 3- Mathematical Morphology- Binary

http://slidepdf.com/reader/full/dip-chapter-3-mathematical-morphology-binary 1/8

1/21/20

Chapter 3 :Mathematical morphologyPart 1 : binary morphology

Dr. Hojeij youssef 

Digital image processing

1

Binary imageBasicDilationErosionOpenCloseNeighbor transform Thin and thickGeodesic reconstruction

Morphology in gray level

Outline

Dr. Hojeij youssef 2

Page 2: DIP Chapter 3- Mathematical Morphology- Binary

7/29/2019 DIP Chapter 3- Mathematical Morphology- Binary

http://slidepdf.com/reader/full/dip-chapter-3-mathematical-morphology-binary 2/8

1/21/20

Binary images have 1 bit per pixel :1 =object pixel0 = non-object pixel (background pixel)

Binary images

Dr. Hojeij youssef 3

Data

Image

Object =real world thing

Component =blob of connectedpixels

Objectsaredescribedbyacollectionof points,whicharedenotedbyvectors.

A set A in a binary image consists of the points α that sharea commonproperty(value)

 Thebackgroundof A is given by thecomplement,

Binary images

Dr. Hojeij youssef 4

1)( A \  A   

CAˆ  A \  Ac   

c A

Page 3: DIP Chapter 3- Mathematical Morphology- Binary

7/29/2019 DIP Chapter 3- Mathematical Morphology- Binary

http://slidepdf.com/reader/full/dip-chapter-3-mathematical-morphology-binary 3/8

1/21/20

Definition : aregion isaset of pixels, whereeachpixel canbereached fromany other pixel in theregion by a finitenumber of steps, whereeach stepstartsatapixel andendsintheneighborhoodof thepixel.

 Twobasicneighborhoods(connectivity)onasquaregrid,denotedby:

N 4 

N 8 

 The paradox of the square

Dr. Hojeij youssef 5

Symmetric around

the origin.

The fundamental operations are :

Complement:

Union:

Intersection:

 Translation:

Fundamental operations

Dr. Hojeij youssef 6

Set algebra :ComplementUnion Intersection…

Boolean algebra :NotOrAndXor

Page 4: DIP Chapter 3- Mathematical Morphology- Binary

7/29/2019 DIP Chapter 3- Mathematical Morphology- Binary

http://slidepdf.com/reader/full/dip-chapter-3-mathematical-morphology-binary 4/8

1/21/20

Dilation

Dr. Hojeij youssef 7

Dilation of set A by structuring element B :

g(x,y)=Or[W{ f(x,y)} ]= dilate(f,W) 

Dilation

Dr. Hojeij youssef 8

Binary dilation effects : Expands the size of 1-valued objects Smoothes object boundaries Closes holes and gapes

Original imageDilatation with

3×3 structuringelement

Dilatation with

7×7 structuringelement

Page 5: DIP Chapter 3- Mathematical Morphology- Binary

7/29/2019 DIP Chapter 3- Mathematical Morphology- Binary

http://slidepdf.com/reader/full/dip-chapter-3-mathematical-morphology-binary 5/8

1/21/20

Erosion

Dr. Hojeij youssef 9

Erosion of setA by structuring element B :

g(x,y)=And[W { f(x,y)} ]= erode(f,W ) 

Erosion

Dr. Hojeij youssef 10

Erosion effects: Shrinks the size of 1-valued objects Smoothes object boundaries Remove peninsulas, and small objects

Originalimage

Erosion by

11×11structuringelement

Erosion by

27×27structuringelement

Erosion by

21×21structuringelement

Page 6: DIP Chapter 3- Mathematical Morphology- Binary

7/29/2019 DIP Chapter 3- Mathematical Morphology- Binary

http://slidepdf.com/reader/full/dip-chapter-3-mathematical-morphology-binary 6/8

1/21/20

Properties

Dr. Hojeij youssef 11

Erosion is dilation of the background

Erosion is not the inverse of dilation

Commutative :

Non-commutative :

Associative:

 Translation invariance :

Properties

Dr. Hojeij youssef 12

Dilation: distributive over the union

Erosion: distributive over the intersection

Multiple dilations:

Page 7: DIP Chapter 3- Mathematical Morphology- Binary

7/29/2019 DIP Chapter 3- Mathematical Morphology- Binary

http://slidepdf.com/reader/full/dip-chapter-3-mathematical-morphology-binary 7/8

1/21/20

Opening and Closing

Dr. Hojeij youssef 13

Openingconsists of a number of erosions followed by the samenumber of dilations :Applications :

Separationof connectedobjectsRemoving small objects (sizediscrimination)

Closingconsists of a number of dilations followed by the samenumber of erosions :Applications :

Connection of disconnected object partsClosingof holesContour smoothing

Rem: close-open and open closeareduals, but not inverses of each other

Majority filter

Dr. Hojeij youssef 14

Binary majority filter is aspecial case of a gray level median filter :g(x,y)=Maj[W{ f(x,y)} ]= majority(f,W) 

Effects : Does not generally shrink or expand objects Smoothes object boundaries Removes small peninsulas, bays, small objects, and small holes Less biased than close-open or open close

Binary imagewith 5%«salt&pepper » noise

Binary imagewith 20%«salt&pepper » noise

3×3 majority filter  3×3 majority filter

Page 8: DIP Chapter 3- Mathematical Morphology- Binary

7/29/2019 DIP Chapter 3- Mathematical Morphology- Binary

http://slidepdf.com/reader/full/dip-chapter-3-mathematical-morphology-binary 8/8

1/21/20

Morphological edge detection

Dr. Hojeij youssef 15

Inner boundary extraction :

Outer boundary extraction :) N  A( A) A( IB

 

) N  A( A) A(OBc

Neighbor transform

Dr. Hojeij youssef 16

Neighbor transform: if v is the element :

 Thin an thick :

Skeleton :

otherwise0

vv / vvsi1) x(v X 

1 x

i

)v X ( X v X   

)v X ( X v X