color a* b* brightness l* texture original image features feature combination e d 22 boundary...

25
Colo r a* b* Brightn ess L* Texture Original Image Features Feature combination E D 2 Boundary Processing Textons A B C A B C 2 Region Processing

Post on 21-Dec-2015

217 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Color a* b* Brightness L* Texture Original Image Features Feature combination E D 22 Boundary Processing Textons A B C A B C 22 Region Processing

Color

a*b*

Brightness

L*

TextureOriginal Image

Features

Feature combination

ED

2

Boundary Processing

Textons

A

B

C

A

B

C

2

Region Processing

Page 2: Color a* b* Brightness L* Texture Original Image Features Feature combination E D 22 Boundary Processing Textons A B C A B C 22 Region Processing

RegionTexture : Histogram of oriented edge response

Boundary and Edge: Edge detection-> lines

Interests points: Corner detection

Page 3: Color a* b* Brightness L* Texture Original Image Features Feature combination E D 22 Boundary Processing Textons A B C A B C 22 Region Processing

Texture Feature (preview)

Texture Gradient TG(x,y,r,) 2 difference of texton histograms• Textons are vector-quantized filter outputs

TextonMap

Page 4: Color a* b* Brightness L* Texture Original Image Features Feature combination E D 22 Boundary Processing Textons A B C A B C 22 Region Processing

Texture boundary

Canny 2MM Us HumanImage

Page 5: Color a* b* Brightness L* Texture Original Image Features Feature combination E D 22 Boundary Processing Textons A B C A B C 22 Region Processing

Texture boundary

Canny 2MM Us HumanImage

Page 6: Color a* b* Brightness L* Texture Original Image Features Feature combination E D 22 Boundary Processing Textons A B C A B C 22 Region Processing

Texture boundary

Canny 2MM Us HumanImage

Page 7: Color a* b* Brightness L* Texture Original Image Features Feature combination E D 22 Boundary Processing Textons A B C A B C 22 Region Processing

Corner Detections (preview)

Page 8: Color a* b* Brightness L* Texture Original Image Features Feature combination E D 22 Boundary Processing Textons A B C A B C 22 Region Processing

Matching with Features

•Detect feature points in both images

•Find corresponding pairs

Page 9: Color a* b* Brightness L* Texture Original Image Features Feature combination E D 22 Boundary Processing Textons A B C A B C 22 Region Processing

Matching with Features

•Detect feature points in both images

•Find corresponding pairs

•Use these pairs to align images

Page 10: Color a* b* Brightness L* Texture Original Image Features Feature combination E D 22 Boundary Processing Textons A B C A B C 22 Region Processing

Boundary and Edge: Edge detection-> lines

Page 11: Color a* b* Brightness L* Texture Original Image Features Feature combination E D 22 Boundary Processing Textons A B C A B C 22 Region Processing

An example:

S.F. in fog S.F. in Canny

Page 12: Color a* b* Brightness L* Texture Original Image Features Feature combination E D 22 Boundary Processing Textons A B C A B C 22 Region Processing

An example:

S.F. in fog S.F. with Hough lines

Page 13: Color a* b* Brightness L* Texture Original Image Features Feature combination E D 22 Boundary Processing Textons A B C A B C 22 Region Processing

Hough Transform

• image edges needs to be grouped into lines and junctions• Hough transform: Detect lines in an edge image

Page 14: Color a* b* Brightness L* Texture Original Image Features Feature combination E D 22 Boundary Processing Textons A B C A B C 22 Region Processing

Line Representation

• is the distance from the origin to the line

is the norm direction of the line

• Image space : Hough space :•

•point in image space ==> a curve in hough space

Page 15: Color a* b* Brightness L* Texture Original Image Features Feature combination E D 22 Boundary Processing Textons A B C A B C 22 Region Processing

Line Representation

• is the distance from the origin to the line

is the norm direction of the line

• Image space : Hough space :•

•point in image space ==> a curve in hough space

For every theta, set:

Page 16: Color a* b* Brightness L* Texture Original Image Features Feature combination E D 22 Boundary Processing Textons A B C A B C 22 Region Processing

Hough Space

• point in hough space ==> line in image space

Page 17: Color a* b* Brightness L* Texture Original Image Features Feature combination E D 22 Boundary Processing Textons A B C A B C 22 Region Processing

Intersection of the curves

• Each pixel in the image => One curve in Hough space

• What is the intersection of the curves?

Page 18: Color a* b* Brightness L* Texture Original Image Features Feature combination E D 22 Boundary Processing Textons A B C A B C 22 Region Processing

Hough Transform

• Points in the line : • In hough space, all the curves pass:• So the intersection of the curves is the parameters of the line! • Next question:

How to find the intersection ?

Page 19: Color a* b* Brightness L* Texture Original Image Features Feature combination E D 22 Boundary Processing Textons A B C A B C 22 Region Processing

Voting Scheme

• Each edge pixel in the image votes in Hough space fora series of

• Choose the of maximum votes

Page 20: Color a* b* Brightness L* Texture Original Image Features Feature combination E D 22 Boundary Processing Textons A B C A B C 22 Region Processing

Basic Hough Transform

Page 21: Color a* b* Brightness L* Texture Original Image Features Feature combination E D 22 Boundary Processing Textons A B C A B C 22 Region Processing

Example

Page 22: Color a* b* Brightness L* Texture Original Image Features Feature combination E D 22 Boundary Processing Textons A B C A B C 22 Region Processing

Example

Page 23: Color a* b* Brightness L* Texture Original Image Features Feature combination E D 22 Boundary Processing Textons A B C A B C 22 Region Processing

Extension

• Choose the sampling of

• Use gradient of the image voting for specific

• Iteratively find the maximum votes and remove

corresponding edge pixels

• Suppress edge pixels close to the detected lines

Page 24: Color a* b* Brightness L* Texture Original Image Features Feature combination E D 22 Boundary Processing Textons A B C A B C 22 Region Processing

Example of Using Estimated Edge Orientation+Iterative line removal

Page 25: Color a* b* Brightness L* Texture Original Image Features Feature combination E D 22 Boundary Processing Textons A B C A B C 22 Region Processing