Introduction to Related Papers of Vessel Segmentation Methods
Advisor : Ku-Yaw Chang
Student : Wei-Lu Lin
2015/1/7
2 Outline
Introduction Related Papers Conclusion
2015/1/7
3 Introduction
What Is Segmentation ?
2015/1/7
4 Introduction
What Is Segmentation ?
2015/1/7
5 Introduction
What Is Segmentation ?
2015/1/7
6 Introduction
Applications Medical Imaging(v)
Object Detection
Recognition Tasks
Traffic Control Systems
Video Surveillance
2015/1/7
People Detection[1] License Plate Recognition[2]
Vessel Segmentation
7 Introduction
Vessel Segmentation Classification Pattern Recognition Techniques
Model-based
Tracking-based
Artificial Intelligence-based
Neural Network-based
Miscellaneous Tube-like Object Detection
2015/1/7
8 Introduction
Vessel Segmentation Classification Pattern Recognition Techniques
Model-based
Tracking-based
Artificial Intelligence-based
Neural Network-based
Miscellaneous Tube-like Object Detection
2015/1/7
9 Introduction
Pattern Recognition Techniques Automatic Detection
Classification
Features
Disadvantage Be Difficult to Deal with Edge Noises and Branch Vessels.
2015/1/7
10 Related Papers – Adaptive Segmentation of Vessels from Coronary
Angiograms Using Multi-scale Filtering
Based on Pattern Recognition Techniques Classification
Steps Select Well-contrast Angiograms
Vessels Segmentation from the Well-contrast Angiograms Using Multi-scale Hessian Matrix
2015/1/7
11
Results
2015/1/7
Related Papers -Adaptive Segmentation of Vessels from Coronary
Angiograms Using Multi-scale Filtering
12 Introduction
Vessel Segmentation Classification Pattern Recognition Techniques
Model-based
Tracking-based
Artificial Intelligence-based
Neural Network-based
Miscellaneous Tube-like Object Detection
2015/1/7
13 Introduction
Model-based Deformable Models
Parametric Models
Template Matching
Generalized Cylinders
Disadvantage Be Hard to Set Model Parameters and Affect the Computational
Cost
2015/1/7
14
Based on Model-based Classification Steps
Initialize Location and Contour
Local Morphology Fitting(LMF) Growing
2015/1/7
Related Papers -Local Morphology Fitting Active Contour for Automatic
Vascular Segmentation
15
Results
2015/1/7
Related Papers -Local Morphology Fitting Active Contour for Automatic
Vascular Segmentation
16 Introduction
Vessel Segmentation Classification Pattern Recognition Techniques
Model-based
Tracking-based
Artificial Intelligence-based
Neural Network-based
Miscellaneous Tube-like Object Detection
2015/1/7
17 Introduction
Tracking-based Manual Start Points
Local Operators
Focus Known to Be a Vessel and Track It
Disadvantage Cannot Effectively Track Vessels in Complex Background
Mostly Rely on the Manual Setting
2015/1/7
18
Based on Tracking-based Classification Steps
Automatic Identification of Start Points
Tracking Based on Bayesian
2015/1/7
Related Papers -A Retinal Vessel Tracking Method
Based On Bayesian Theory
19
Results
2015/1/7
Related Papers -A Retinal Vessel Tracking Method
Based On Bayesian Theory
other this paper
20 Conclusion
Segmentation Algorithms A lot of methods
Future Persuade more faster, more accurate and more automated
In My Opinion Automation is not important
Interaction
2015/1/7
21 Conclusion
2015/1/7
22 Conclusion
2015/1/7
Contrast Image Image Contrast Image Image
Contrast Image Image Contrast Image Image
23 References
[1http://www.di.ens.fr/~laptev/objectdetection.html [2] http://bit.ly/1BNIAem https://www.youtube.com/watch?v=ceIddPk78yA&list
=PLz8K9D6W9hwa1I-LZhmyC7ux1fXi1VSzu&index=8
https://www.youtube.com/watch?v=EOCqzA2luy0
https://www.youtube.com/watch?v=lwMuJX480jo
2015/1/7
24
The EndThank you for listening
2015/1/7