18-09-2016, 04:15 PM
19-09-2016, 10:45 AM
Biometrics are automated methods of recognizing a person based on a physiological or behavioral characteristics (fingerprint, iris, voice recognition etc.) Among them one of the most
effective biometric technology is the blood vessel recognition of ocular fundus, especially the retina. Personal identification using retinal scans is composed of three principal modules: blood
vessel segmentation, feature extraction, and feature matching. Due to the technological limitations in manufacturing low-cost retinal scanners, there were few works published on
human retina identification, and algorithms for analyzing vessel tree patterns for biometric tasks are currently under intensive development.
Objective:
To implement procedures for blood vessel segmentation of retinal images in MATLAB for biometric application. The final result of this work (amount will differ for BA and MA) should
comprise:
A comparison of existing state-of-the-art retinal image processing techniques;
A working MATLAB program which is able to compare retinal images for personal identification using existing image processing techniques;
A conclusion on advantages and disadvantages of each implemented image processing technique.
Tasks involved:
To investigate current blood vessel extraction techniques in retinal images and vascular models in general;
To learn image processing in MATLAB;
To make an algorithm for blood vessel segmentation;
To validate the algorithm by using publicly available databases of retinal images.
Requirements:
High self-motivation
Interest in image processing techniques
Participation in „Machine Vision“ course from MRT (KIT)
Experience in MATLAB is desirable
We provide:
Supervisor who will supervise
Opportunity to practice image processing skills in on-the-edge-of-science application
The possibility to develop and pursue your own ideas within the project.
effective biometric technology is the blood vessel recognition of ocular fundus, especially the retina. Personal identification using retinal scans is composed of three principal modules: blood
vessel segmentation, feature extraction, and feature matching. Due to the technological limitations in manufacturing low-cost retinal scanners, there were few works published on
human retina identification, and algorithms for analyzing vessel tree patterns for biometric tasks are currently under intensive development.
Objective:
To implement procedures for blood vessel segmentation of retinal images in MATLAB for biometric application. The final result of this work (amount will differ for BA and MA) should
comprise:
A comparison of existing state-of-the-art retinal image processing techniques;
A working MATLAB program which is able to compare retinal images for personal identification using existing image processing techniques;
A conclusion on advantages and disadvantages of each implemented image processing technique.
Tasks involved:
To investigate current blood vessel extraction techniques in retinal images and vascular models in general;
To learn image processing in MATLAB;
To make an algorithm for blood vessel segmentation;
To validate the algorithm by using publicly available databases of retinal images.
Requirements:
High self-motivation
Interest in image processing techniques
Participation in „Machine Vision“ course from MRT (KIT)
Experience in MATLAB is desirable
We provide:
Supervisor who will supervise
Opportunity to practice image processing skills in on-the-edge-of-science application
The possibility to develop and pursue your own ideas within the project.