CSE4227 Digital Image Processing

Chapter 10 – Image Segmentation Part II

(Region based)

Dr. Kazi A Kalpoma

Professor, Department of CSE
Ahsanullah University of Science & Technology (AUST)

Contact: [email protected]
Google Class code: fmxdafw

CSE | AUST Spring 2024

Today’s Contents
❑Region Based Segmentation
❑What is a Region
❑Region Growing, and
❑Region Splitting & Merging
❑Thresholding Based Segmentation
❑GLOBAL
❑ADAPTIVE
❑LOCAL

• Chapter 10 from R.C. Gonzalez and R.E. Woods, Digital Image Processing
(3rd Edition), Prentice Hall, 2008 [ Section 10.1, 10.2 (excluding 10.2.7) ]
 Segmentation Algorithms
Segmentation can be performed based on one of the
following two basic properties of intensity values:

❑Similarity
⮚Partitioning an image into regions that are
similar according to a set of predefined
criteria.
❑Discontinuity
⮚Detecting boundaries of regions based on
local discontinuity in intensity.
Region based segmentation

• In edge detection we segment an image by

identifying the boundaries of the objects.
• These boundaries are the locations where the
intensity is changed.
• In the region based segmentation approach,
we will identify regions occupied by the
objects.
• We will group pixels which are similar in some
region property.
Edge Pixels
20 0 0 0 0 0 0 0

20 0 0 0 0 0 0 0

20 0 143 143 143 143 0 0

20 0 143 200 200 143 0 0

20 0 143 200 200 143 0 0

20 0 143 143 143 143 0 0

20 0 0 0 0 0 0 0

20 0 0 0 0 0 0 0

8x8 cropped Image

Edge based vs Region based Segmentation
20 0 0 0 0 0 10 10

20 0 0 0 0 0 10 10

20 0 143 143 143 143 10 10

20 0 143 200 200 143 10 10

20 0 143 200 200 143 10 10

20 0 143 143 143 143 10 10

20 0 0 0 0 0 10 10

20 0 0 0 0 0 10 10

8x8 cropped Image

Edge based vs Region based Segmentation
20 0 0 0 0 0 10 10

20 0 0 0 0 0 10 10

20 0 143 143 143 143 10 10

20 0 143 200 200 143 10 10

20 0 143 200 200 143 10 10

20 0 143 143 143 143 10 10

20 0 0 0 0 0 10 10

20 0 0 0 0 0 10 10
Region Based Segmentation Method

❑Segments the image into various regions having

similar characteristics.

❑The two basic techniques based on this method

are:
i) Region Growing, and
ii) Region Splitting & Merging.
Region Growing
Region growing techniques start with one pixel of a
potential region and try to grow it by adding adjacent
pixels till the pixels being compared are too disimilar.

• The first pixel selection can be just the first unlabeled

pixel in the image or a set of seed pixels can be chosen
from the image.

• Usually a statistical test is used to decide which pixels

can be added to a region.

12
 Region-Based Segmentation

Example: Region Growing based on 8-connectivity

 Region Growing Algorithm

* 14
Example

4-connectivity
Example

4-connectivity
Example

8-connectivity
Example
 Region Splitting and Merging

❑ Let R represent the entire spatial region occupied by an

image. Image segmentation is a process that partitions R
into n sub-regions, R1, R2, …, Rn, such that

Logical predicate
Region Splitting and Merging
■ Region Splitting
❑ Region Growing: Starts from a set of seed points.
❑ Region Splitting: Starts with the whole image as a single
region and subdivide the regions that do not satisfy a
condition.
❑ Image = One Region R
❑ Select a predicate P (gray values etc.)
❑ Successively divide each region into smaller and smaller
quadrant regions so that:

* Image Segmentation 20
Region Splitting and Merging
■ Region Splitting

Problem? Adjacent regions could be same

Solution? Allow Merge

* Image Segmentation 21
Region-Based Segmentation
■ Region Merging
❑ Region merging is the opposite of region splitting.
❑ Merge adjacent regions Ri and Rj for which:

■ Region Splitting/Merging
❑ Stop when no further split or merge is possible

* Image Segmentation 22
Region-Based Segmentation
■ Example

1. Split into four disjointed quadrants any region Ri where P(Ri)=False

2. Merge any adjacent regions Rj and Rk for which P(Rj U Rk)=True

3. Stop when no further merging or splitting is possible

* Image Segmentation
23
Region Splitting and Merging Example
Split into four disjoint quadrant and select a threshold first…say T <= 4.

5 6 6 6 7 7 6 6

6 7 6 7 5 5 4 7

6 6 4 4 3 2 5 6

5 4 5 4 2 3 4 6

0 3 2 3 3 2 4 7

0 0 0 0 2 2 5 6

1 1 0 1 0 3 4 4

1 0 1 0 2 3 5 4

Please see the video recording to understand the example.

* Link is provided onSegmentation
Image the google classroom. 24
 Region Splitting Example
Split into quadrant if T < (max-min), continue until no more spit is possible

5 6 6 6 7 7 6 6

6 7 6 7 5 5 4 7

6 6 4 4 3 2 5 6

5 4 5 4 2 3 4 6

0 3 2 3 3 2 4 7

0 0 0 0 2 2 5 6

1 1 0 1 0 3 4 4

1 0 1 0 2 3 5 4

Please see the video recording to understand the example.

* Link is provided onSegmentation
Image the google classroom. 25
 Region Splitting Example
Split into quadrant if T < (max-min), continue until no more spit is possible

5 6 6 6 7 7 6 6

6 7 6 7 5 5 4 7

6 6 4 4 3 2 5 6

5 4 5 4 2 3 4 6

0 3 2 3 3 2 4 7

0 0 0 0 2 2 5 6

1 1 0 1 0 3 4 4

1 0 1 0 2 3 5 4

Please see the video recording to understand the example.

* Link is provided onSegmentation
Image the google classroom. 26
 Region Splitting Example
Now terminate splitting and go for Merge the adjacent regions If T is greater

5 6 6 6 7 7 6 6

6 7 6 7 5 5 4 7

6 6 4 4 3 2 5 6

5 4 5 4 2 3 4 6

0 3 2 3 3 2 4 7

0 0 0 0 2 2 5 6

1 1 0 1 0 3 4 4

1 0 1 0 2 3 5 4

Please see the video recording to understand the example.

* Link is provided onSegmentation
Image the google classroom. 27
 Region Merging Example
Merge the adjacent regions If T is greater than (maxi-minj)
AND (maxj-mini) of both regions

5 6 6 6 7 7 6 6

6 7 6 7 5 5 4 7

6 6 4 4 3 2 5 6

5 4 5 4 2 3 4 6

0 3 2 3 3 2 4 7

0 0 0 0 2 2 5 6

1 1 0 1 0 3 4 4

1 0 1 0 2 3 5 4

Please see the video recording to understand the example.

* Image Segmentation 28
Link is provided on the google classroom.
 Region
RegionMerging
MergingExample….
Example
Merge the adjacent regions If T is greater than (maxi-minj)
AND (maxj-mini) of both regions

5 6 6 6 7 7 6 6

6 7 6 7 5 5 4 7

6 6 4 4 3 2 5 6

5 4 5 4 2 3 4 6

0 3 2 3 3 2 4 7

0 0 0 0 2 2 5 6

1 1 0 1 0 3 4 4

1 0 1 0 2 3 5 4

Please see the video recording to understand the example.

* Image Segmentation 29
Link is provided on the google classroom.
 Region Merging Example….
Merge the adjacent regions If T is greater than (maxi-minj)
AND (maxj-mini) of both regions

5 6 6 6 7 7 6 6

6 7 6 7 5 5 4 7

6 6 4 4 3 2 5 6

5 4 5 4 2 3 4 6

0 3 2 3 3 2 4 7

0 0 0 0 2 2 5 6

1 1 0 1 0 3 4 4

1 0 1 0 2 3 5 4

Please see the video recording to understand the example.

* Image Segmentation 30
Link is provided on the google classroom.
 Region Merging Example….
Merge the adjacent regions If T is greater than (maxi-minj)
AND (maxj-mini) of both regions

5 6 6 6 7 7 6 6

6 7 6 7 5 5 4 7

6 6 4 4 3 2 5 6

5 4 5 4 2 3 4 6

0 3 2 3 3 2 4 7

0 0 0 0 2 2 5 6

1 1 0 1 0 3 4 4

1 0 1 0 2 3 5 4

Please see the video recording to understand the example.

* Image Segmentation 31
Link is provided on the google classroom.
 Region Merging Example….
Merge the adjacent regions If T is greater than (maxi-minj)
AND (maxj-mini) of both regions

5 6 6 6 7 7 6 6

6 7 6 7 5 5 4 7

6 6 4 4 3 2 5 6

5 4 5 4 2 3 4 6

0 3 2 3 3 2 4 7

0 0 0 0 2 2 5 6

1 1 0 1 0 3 4 4

1 0 1 0 2 3 5 4

Please see the video recording to understand the example.

* Image Segmentation 32
Link is provided on the google classroom.
 Region Merging Example….
Merge the adjacent regions If T is greater than (maxi-minj)
AND (maxj-mini) of both regions

5 6 6 6 7 7 6 6

6 7 6 7 5 5 4 7

6 6 4 4 3 2 5 6

5 4 5 4 2 3 4 6

0 3 2 3 3 2 4 7

0 0 0 0 2 2 5 6

1 1 0 1 0 3 4 4

1 0 1 0 2 3 5 4

Please see the video recording to understand the example.

* Image Segmentation 33
Link is provided on the google classroom.
 Region Merging Example….
Merge the adjacent regions If T is greater than (maxi-minj)
AND (maxj-mini) of both regions

5 6 6 6 7 7 6 6

6 7 6 7 5 5 4 7

6 6 4 4 3 2 5 6

5 4 5 4 2 3 4 6

0 3 2 3 3 2 4 7

0 0 0 0 2 2 5 6

1 1 0 1 0 3 4 4

1 0 1 0 2 3 5 4

Please see the video recording to understand the example.

* Image Segmentation 34
Link is provided on the google classroom.
 Region Merging Example….
Merge the adjacent regions If T is greater than (maxi-minj)
AND (maxj-mini) of both regions

5 6 6 6 7 7 6 6

6 7 6 7 5 5 4 7

6 6 4 4 3 2 5 6

5 4 5 4 2 3 4 6

0 3 2 3 3 2 4 7

0 0 0 0 2 2 5 6

1 1 0 1 0 3 4 4

1 0 1 0 2 3 5 4

Please see the video recording to understand the example.

* Image Segmentation 35
Link is provided on the google classroom.
 Region Merging Example….
Merge the adjacent regions If T is greater than (maxi-minj)
AND (maxj-mini) of both regions

5 6 6 6 7 7 6 6

6 7 6 7 5 5 4 7

6 6 4 4 3 2 5 6

5 4 5 4 2 3 4 6

0 3 2 3 3 2 4 7

0 0 0 0 2 2 5 6

1 1 0 1 0 3 4 4

1 0 1 0 2 3 5 4

Please see the video recording to understand the example.

* Image Segmentation 36
Link is provided on the google classroom.
 https://www.youtube.com/watch?v=hKYBNdSspIo

* Image Segmentation 37
Thresholding
■ Segmentation into two classes/groups
❑ Foreground (Objects)
❑ Background

* Image Segmentation 38
Thresholding

* Image Segmentation 39
Thresholding

Objects & Background

1.0 r > threshold

s=
0.0 r <= threshold

* Image Segmentation 40
Thresholding

❖ GLOBAL
❖ LOCAL / REGIONAL
❖ ADAPTIVE

* Image Segmentation 41
Thresholding

* Image Segmentation 42
Thresholding

* Image Segmentation 43
Global Thresholding
■ Single threshold value for entire image
■ Fixed ?
■ Automatic
❑ Intensity histogram

* Image Segmentation 44
Global Thresholding
■ Single threshold value for entire image
■ Fixed ?
■ Automatic
❑ Intensity histogram

* Image Segmentation 45
Global Thresholding

* Image Segmentation 46
 https://www.youtube.com/watch?v=f1SaYzOthCM

* Image Segmentation 47
Global Thresholding?

Multilevel Thresholding can be a solution?

* Image Segmentation 48
Multilevel Thresholding

* Image Segmentation 49
Multilevel Thresholding

* Image Segmentation 50
Multilevel Global Thresholding?

* Image Segmentation 51
Basic Adaptive Thresholding

* Image Segmentation 52
Local Adaptive Thresholding
❑The Local Adaptive Thresholding chooses
different threshold values for every pixel in
the image based on an analysis of its neighboring
pixels.
■ Threshold: function of neighboring pixels

Image Segmentation 53
Local Adaptive Thresholding
■ Niblack Algorithm

■ Neighborhood size???

* Image Segmentation 54
Thresholding

* Image Segmentation 55
https://youtu.be/nPyb2BVBeB8

* Image Segmentation 56
K-Means Clustering
1. Chose the number (K) of clusters and randomly select the
centroids of each cluster.
2. For each data point:
❑ Calculate the distance from the data point to each cluster.
❑ Assign the data point to the closest cluster.
3. Recompute the centroid of each cluster.
4. Repeat steps 2 and 3 until there is no further change in
the assignment of data points (or in the centroids).

* Image Segmentation 57