1. Instructor: Dr. G. Bebis Reza Amayeh Fall 2005
Anil K. Jain, Arun Ross and Sharath Pankanti, “A Prototype Hand Geometry-based Verification System” Slobodan Ribaric and Ivan Fratric, “A Biometric Identification System Based on Eigen-palm and Eigen-finger Features”
Instructor: Dr. G. Bebis
Reza Amayeh
Fall 2005

2. Why Hand Geometry Each biometrics has its strengths and limitation
What is the most effective biometric measurement?
Each biometrics has its strengths and limitation
It’s depend on application
User acceptability is the most significant factor
For many access control applications, like dormitory meal plan access, very distinctive biometrics, e.g., fingerprint and iris, may not be acceptable
Hand geometry-based verification systems are not new and have been available since the early 1970s

3. A Prototype Hand Geometry-based Verification System
A technique to measure of hand geometry is introduced by using probability function
Outines:
Image Acquisition
Feature Extraction
Verification Phase

4. Image Acquisition
The image acquisition system comprises of a light source, a camera, a single mirror and a surface (with five pegs on it)
Pegs serve as control points for appropriate placement of the right hand of the user
A 640 by bit grayscale image of the hand is captured

5. Feature Extraction
Typically features include length and width of the fingers, aspect ratio of the palm or fingers, thickness of the hand, etc.
16 axes along which the various features mentioned above have been measured
The hand is represented as a vector of the measurements selected above

6. Feature Extraction is the gray values of the pixels along a axis
A sequence of pixels along a measurement axis will not have an ideal gray scale profile
Background lighting, color of the skin, noise and etc.

7. Feature Extraction Modeling profile: Assumption:
The observed profile is obtained from the ideal profile by the addition of Gaussian noise to each of the pixels
The gray level of an arbitrary pixel along a particular axis is independent of the gray level of other pixels in the line
Operating under these assumptions, we can write the joint distribution of all the pixel values along a particular axis as

8. Feature Extraction
The goal now is to estimate using the observed pixel values along the chosen axis
Use the Maximum Likelihood Estimate (MLE) method to estimate parameters q
likelihood function:

9. Feature Extraction The parameters can now be estimated iteratively
The initial estimates of parameters are obtained as follows:
are estimated using the gray values of the first pixels along the axis
are estimated using the gray values of the pixels from
are estimated using the gray values of the pixels between
The values of are fixed for the system

10. Feature Extraction
Result:

11. Verification Phase
Let represent the d-dimensional feature vector in the database associated with the claimed identity and be the feature vector of the hand whose identity has to be verified.

12. Exp. Result
The hand geometry authentication system was trained and tested using a database of 50 users
Ten images of each user's hand were captured over two sessions
In each session the background lighting of the acquisition device was changed
Thus a total of 500 images were made available

13. Questions?

14. A Biometric Identification System Based on Eigen-palm and Eigen-finger Features
The identification process can be divided into the following phases:
Capturing the image
Preprocessing, extracting and normalizing the palm and strip-like finger subimages
Extracting the eigenpalm and eigenfinger features based on the PCA
Matching and fusion
A decision based on the (k, l)-NN classifier and thresholding

15. System Description

16. Preprocessing and Sub-images Extraction
(a) Example of a scanned hand image. (b) Binerized hand image.

17. Preprocessing and Sub-images Extraction
(a) The hand contour and the relevant points for finding the regions of interest. (b) Processed finger on the hand contour.

18. Geometry and Lighting Normalization
(a) Original hand image with the regions of interest marked on it. (b) Palm subimage. (c) Little-finger subimage. (d) Ring-finger subimage.
(e) Middle-finger subimage. (f) Index-finger subimage. (g) Thumb subimage.

19. Geometry and Lighting Normalization
After the subimages have been extracted, a lighting normalization using histogram fitting is applied.
In this process, a target histogram G(l) is selected for each of the six subimage
Subimages (a) before and (b) after the histogram fitting.

20. Lighting Normalization
Palm subimages with the corresponding histograms (a) before and (b) after the histogram fitting.

21. Extraction of the Eigenpalm and Eigenfinger Features
Average palm and finger images from database.
Eigenpalms and eigenfingers obtained from the database (indicated by the corresponding ordinal numbers): (a) the eigenpalms and (b) the
little-finger eigenfingers.

22. MATCHING AND FUSION AT THE MATCHING-SCORE LEVEL
The matching between corresponding feature vectors is based on the Euclidean distance
The distances are normalized and transformed into similarity measures
The frequency distribution of the distances of templates of the same person is calculated.

23. MATCHING AND FUSION AT THE MATCHING-SCORE LEVEL
The normalized outputs of the six matching modules are combined using fusion at the matching-score level.

24. Experimental Results (Recognition)
110 People and 5 templates per person.
Minimum Euclidean distance.

25. Experimental Results (Recognition)
Recognition results as a function of the subspace dimensionality for (a) palm subspace and (b) finger subspaces.

26. Experimental Results (Identification I )
57 people (10 TPP) as clients and 70 people (10 TPP) as imposter
The first 7 templates as a train set for each client

27. Experimental Results (Identification II )
127 people and 10 templates per person