1. Image Segmentation

2. Concepts and Approaches
What is Image Segmentation?
Image Segmentation Methods
Thresholding
Boundary-based
Region-based: region growing, splitting and merging
Partition an image into regions, each associated with an object
but what defines an object?

3. Thresholding Method thresholding histogram multiple thresholds
single threshold
multiple thresholds

4. Thresholding Method
Global Thresholding: When does It Work?

5. Thresholding Method Global Thresholding: When does It NOT Work?
A meaningful global threshold may not exist
Image-dependent
global
thresholding

6. Thresholding Method true object boundary Thresholding T = 4.5

7. Thresholding Method Solution Spatially adaptive thresholding
Localized processing
Split

8. spatially adaptive threshold selection
Thresholding Method
spatially adaptive threshold selection
Thresholding T = 4
Thresholding T = 7
Thresholding T = 4
Thresholding T = 7

9. merge local segmentation results
Thresholding Method
merge local segmentation results
merge
merge
merge
merge

10. Region-Based Segmentation
Edges and thresholds sometimes do not give good results for segmentation.
Region-based segmentation is based on the connectivity of similar pixels in a region.
Each region must be uniform.
Connectivity of the pixels within the region is very important.
There are two main approaches to region-based segmentation: region growing and region splitting.

11. Region-Based Segmentation
Basic Formulation
Let R represent the entire image region.
Segmentation is a process that partitions R into subregions, R1,R2,…,Rn, such that
where P(Rk): a logical predicate defined over the points in set Rk
For example: P(Rk)=TRUE if all pixels in Rk have the same gray level.

12. Region-Based Method: Region Growing Key: similarity measure
Start from a seed, and let it grow (include similar neighborhood)
Key: similarity measure

13. Region-Based Segmentation Region Splitting and Merging
Region splitting is the opposite of region growing.
First there is a large region (possible the entire image).
Then a predicate (measurement) is used to determine if the region is uniform.
If not, then the method requires that the region be split into two regions.
Then each of these two regions is independently tested by the predicate (measurement).
This procedure continues until all resulting regions are uniform

14. Region-Based Method: Split and Merge
The main problem with region splitting is determining where to split a region.
One method to divide a region is to use a quadtree structure.
Quadtree: a tree in which nodes have exactly four descendants

15. Region-Based Segmentation Region Splitting and Merging
The split and merge procedure:
Split into four disjoint quadrants any region Ri for which P(Ri) = FALSE.
Merge any adjacent regions Rj and Rk for which P(RjURk) = TRUE. (the quadtree structure may not be preserved)
Stop when no further merging or splitting is possible.

16. Region-Based Method: Split and Merge
Example: Quadtree Split and Merge Procedure
Iteration 1
split
merge
original image
4 regions
4 regions
(nothing to merge)
Split Step  split every non-uniform region to 4
Merge Step  merge all uniform adjacent regions

17. Region-Based Method: Split and Merge
Example: Quadtree Split and Merge Procedure
Iteration 2
split
merge
from Iteration 1
13 regions
4 regions
Split Step  split every non-uniform region to 4
Merge Step  merge all uniform adjacent regions

18. Region-Based Method: Split and Merge
Example: Quadtree Split and Merge Procedure
Iteration 3
split
merge
from Iteration 2
10 regions
2 regions
Split Step  split every non-uniform region to 4
Merge Step  merge all uniform adjacent regions
final segmentation result

19. Hard Problem: Textures Similarity measure makes the difference