Sanun Srisuk
42973003
1

Outline

Problem Statement

Theory & Algorithm

Results

Conclusion
2

Problem Statement
Segmentation using small scale Segmentation using large scale
3

Theory & Algorithm is a pixel in an image.
is an edge energy at location s along the orientation theta.
is the probability of finding an image boundary in the direction theta from s .
is the probability of finding an image boundary in the direction theta+pi from s .
4

Gaussian Function
5

3-D GD mask
6

GD mask in different theta
7

3-D DOOG mask
8

DOOG mask in different theta
9

Intensity Edges
The edge energy E(s, theta) at scale sigma is defined to be the magnitude of the gradient of the smoothed image
, which is obtained by smoothing the original image I(x,y) with a Gaussian kernel
10

Intensity Edges
P(s, theta) is the probability of finding an image boundary in the direction theta from s .
11

Conventional Edge Detection & EdgeFlow
12

Edge Flow Vector where is a complex number with its magnitude representing the resulting edge energy and angle representing the flow direction.
13

Results red green blue intensity
14

Intensity & Texture EdgeFlow
15

Total EdgeFlow
16

Segmentation Results
17

Segmentation Results
18

EdgeFlow
19

EdgeFlow
20

EdgeFlow Results
21

EdgeFlow Results
22

EdgeFlow Results
23

EdgeFlow Results
24

Segmentation results in different theta
25

Segmentation results in different theta
26

Conclusion

EdgeFlow using a predictive coding model to identify and integrate the direction of change in image attributes such as color, texture, and phase discontinuities, at each image location.
27

The End.
28