REGION
GROWING
ALGORITHM
IMAGE SEGMENTATION
Image segmentation is the process of dividing a digital image
into multiple segments or regions, each representing a
meaningful part of the image.
Types of Segmentation Techniques:
1. Thresholding – Based on pixel intensity
2. Edge-based – Detects boundaries using gradients
3. Region-based – Groups similar pixels (e.g., Region
Growing)
4. Clustering – Like K-means clustering of pixel values
5. Deep Learning – Using CNNs (e.g., U-Net for semantic
segmentation)
REGION BASED SEGMENTATION
Edges and threshold do not give good result for segmentation
so we go for Region based segmentation
Region based segmentation is based on connectivity of similar
pixels in a region
Two approaches:
Region Growing
Region Splitting
REGION GROWING
Region Growing is a region-based image segmentation
technique used to divide an image into meaningful parts
based on pixel similarity (similarity based on texture,
colour, graylevel, shape…)
STEPS FOR THIS APPROACH:
1. Selection of a initial seed point.
2. Seed growing criteria
3. Termination of the segmentation process
For the given image, show the result of region growing algorithm
1 0 7 9 8
0 1 8 8 9
1 1 7 9 9
0 0 9 8 9
2 1 9 8 9
➢ Given here the seed point is 9 ➢ Given here the seed point is 1
➢ Now to find the similarity criteria ➢ Now to find the similarity criteria
The threshold value is T <= 4 The threshold value is T <= 4
| f(x,y)- f(x’-y’) | <= T | f(x,y)- f(x’-y’) | <= T
| f(x,y)- 9| <= 4 | f(x,y)- 1| <= 4
F(x,y)= {5,6,7,8,9} ------> A F(x,y)= {0,1,2,3,4} ------> B
Where the values are in the image it is Where the values are in the image it
considered as one part is considered as one part
RESULTANT IMAGE:
A A B B B
A A B B B
A A B B B
A A B B B
A A B B B
Here A is one segment and B is another segment
EXAMPLE PROBLEM
1 0 7 9 8
0 1 8 8 9
1 1 7 9 9
0 0 9 8 9
2 1 9 8 9
➢ Here the seed point is 9
➢ Now to find the similarity criteria
The threshold value is T < = 4
➢ Now to find difference between the Seed point
and the intensity (pixel) values.
Now to find difference between the Seed point and the intensity
(pixel) values.
Seed point – pixel values
1 0 7 9 8 8 9 2 0 1
0 1 8 8 9 9 8 1 1 0
1 1 7 9 9 8 8 2 0 0
0 0 9 8 9 9 9 0 1 0
2 1 9 8 9 7 8 0 1 0
Now the value below the threshold is to one part and the
other is another part (t<4 is one part segmented)
EXAMPLE PROBLEM 2:
5 6 6 7 6 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
0 0 1 0 2 3 5 4
➢ Given here the seed point is 7
➢ Now to find the similarity criteria
The threshold value is T < = 3
➢ Now to find difference between the Seed
point and the intensity (pixel) values.
Now to find difference between the Seed point and the
intensity (pixel) values.
Seed point – pixel values
5 6 6 7 6 7 6 6 2 1 1 0 1 0 1 1
6 7 6 7 5 5 4 7 1 0 1 0 2 2 3 0
6 6 4 4 3 2 5 6 1 1 3 3 4 5 2 1
5 4 5 4 2 3 4 6 2 3 2 3 4 4 3 1
0 3 2 3 3 2 4 7 7 4 5 4 4 5 3 0
0 0 0 0 2 2 5 6 7 7 7 7 5 5 2 1
1 1 0 1 0 3 4 4 6 6 7 6 7 4 3 3
0 0 1 0 2 3 5 4 7 7 6 7 5 4 2 3
ADVANTAGES
➢ Simple and easy to implement.
➢ Gives accurate region boundaries when intensity values are
clear.
➢ Flexible for different types of images.
DISADVANTAGES
➢ Needs a good starting point and threshold.
➢ Can be slow for large or noisy images.
➢ Not suitable for overlapping or complex regions.
