15
1 Segmentation Using Segmentation Using Texture Texture

1 Segmentation Using Texture. 2 Project Description Input: satellite image and a texture Task: segmentation of the image based on the texture Output:

Embed Size (px)

Citation preview

Page 1: 1 Segmentation Using Texture. 2 Project Description Input: satellite image and a texture Task: segmentation of the image based on the texture Output:

1

Segmentation Using TextureSegmentation Using Texture

Page 2: 1 Segmentation Using Texture. 2 Project Description Input: satellite image and a texture Task: segmentation of the image based on the texture Output:

2

Project DescriptionProject Description

Input: satellite image and a textureTask: segmentation of the image based on

the textureOutput: labeled image

Page 3: 1 Segmentation Using Texture. 2 Project Description Input: satellite image and a texture Task: segmentation of the image based on the texture Output:

3

What Is a Texture ?What Is a Texture ?

There are many definitions of the word texture: Describes something that has a surface that is not

smooth but has a raised pattern on it (from Cambridge advanced learner's dictionary)

A measure of the variation of the intensity of a surface, quantifying properties such as smoothness, coarseness and regularity (from FOLDOC - computing dictionary)

Page 4: 1 Segmentation Using Texture. 2 Project Description Input: satellite image and a texture Task: segmentation of the image based on the texture Output:

4

AlgorithmsAlgorithms

Histogram matchingLaw’s texture measureRun-length matrices

Page 5: 1 Segmentation Using Texture. 2 Project Description Input: satellite image and a texture Task: segmentation of the image based on the texture Output:

5

The basic idea is to compute the histogram of the template, and then sweep a window over the image, compute the histogram of the window and do a correlation between the histograms.

Histogram Matching Algorithm IHistogram Matching Algorithm I

The texture we are searching (the template)Window at step k

(the sample)

Window at step k+1

Short description:

Page 6: 1 Segmentation Using Texture. 2 Project Description Input: satellite image and a texture Task: segmentation of the image based on the texture Output:

6

Histogram Matching Algorithm IIHistogram Matching Algorithm II

i. Histogram equalization (HE) of the image:

ii. Calculate the histogram of the texture

iii. Overlap the image by the texture at each possible position and calculate correlation of the histogram of the texture f and the one of the overlapped area g:

FOR MORE INFO...

Histogram Transformation in Image Processing and Its Applications by Attila Kuba, University of Szeged

x y

nymxgyxfnmR ),(),(),(

Page 7: 1 Segmentation Using Texture. 2 Project Description Input: satellite image and a texture Task: segmentation of the image based on the texture Output:

7

Histogram Matching Algorithm IIIHistogram Matching Algorithm III

iv. Thresholding of the correlation map:i. High correlated values are set to 1

ii. Low correlated values are set to 0

This yields a binary image BI

v. Median filter to eliminate the holes on BI

vi. Border := BI – erosion(BI)

vii. Put the border on the original image

You can choose an algorithm for the search (we have more than one )

You should wait (but not too long) for the resulting image

OBSERVATION...

Page 8: 1 Segmentation Using Texture. 2 Project Description Input: satellite image and a texture Task: segmentation of the image based on the texture Output:

8

Histogram Matching Algorithm IVHistogram Matching Algorithm IV

Zoomed texture

Page 9: 1 Segmentation Using Texture. 2 Project Description Input: satellite image and a texture Task: segmentation of the image based on the texture Output:

9

Histogram Matching Algorithm VHistogram Matching Algorithm V

Zoomed texture

Page 10: 1 Segmentation Using Texture. 2 Project Description Input: satellite image and a texture Task: segmentation of the image based on the texture Output:

10

Run-length Algorithm IRun-length Algorithm I

33333

22222

32222

33222

32222

City – rough grayscale variations – short runs

Grass – smooth grayscale variations – long runs

50022

58300

00000

00000

00005

00003

00002

00005

00004

00006

= P

= P

Page 11: 1 Segmentation Using Texture. 2 Project Description Input: satellite image and a texture Task: segmentation of the image based on the texture Output:

11

Run-length Algorithm IIRun-length Algorithm II

Second step: Calculate short run emphasis Calculate long run emphasis Calculate gray level nonuniformity Find closest matches

Tang, Xiaoou, “Texture Information in Run-Length Matrices”, IEEE transactions on image processing, vol. 7, no 11, november 1998 http://www.s2.chalmers.se/undergraduate/courses0203/ess060/PDFdocuments/ForPrinter/Notes/TextureAnalysis.pdf

FOR MORE INFO...

Page 12: 1 Segmentation Using Texture. 2 Project Description Input: satellite image and a texture Task: segmentation of the image based on the texture Output:

12

Law’s Texture Measure ILaw’s Texture Measure IFirst step:

Original image

Verticalkernel

Horizontal kernel

Measure energy

Measure energy

Law’s energymatrix

Chantler, Michael J, “The effect of variation in illuminant direction on texture classification”, pp 90-134, http://www.cee.hw.ac.uk/~mjc/texture/mjc-phd/

FOR MORE INFO...

Page 13: 1 Segmentation Using Texture. 2 Project Description Input: satellite image and a texture Task: segmentation of the image based on the texture Output:

13

Law’s Texture Measure IILaw’s Texture Measure IISecond step:

Law’s energymatrix

Grayscaledilation Thresholding Binary

dilation

Segmented image

Krabbe, Susanne, “Still Image Segmentation”, http://www-mm.informatik.unimannheim.de/veranstaltungen/animation/multimedia/segmentation/documentation/Segmentation.pdf

FOR MORE INFO...

Page 14: 1 Segmentation Using Texture. 2 Project Description Input: satellite image and a texture Task: segmentation of the image based on the texture Output:

14

Law’s Texture Measure IIILaw’s Texture Measure IIIOriginal image Output image

Page 15: 1 Segmentation Using Texture. 2 Project Description Input: satellite image and a texture Task: segmentation of the image based on the texture Output:

15

Kornél Kovács

Blaž Luin

Zoltán Kiss

Dumitru Şipoş