Palmprint Verification
Acknowledgements
PALMPRINT DENTIFICATION USING EFAUSDORF’F. DISTANCE by Fang Li,
Maylor K.H. Leung, Xaozhou Yu
Real-time palmprint acquisition system design by M. Wong, D. Zhang,
W.-K. Kong and G. Lu.
Discrete Finger and Palmar Feature Extraction for Personal
Authentication by Junta Doi, Member, IEEE, and Masaaki Yamanaka
Online Palmprint Identification by David Zhang, Senior Member, IEEE,
Wai-Kin Kong, Member, IEEE,Jane You, Member, IEEE, and Michael
Wong
Two novel characteristics in palmprint veriÞcation: datum point
invariance and line feature matching by Dapeng Zhang*, Wei Shu
Introduction
Palm Patterns are utilized in many applications:
To correlate palm patterns with medical disorders, e.g. genetic
disorders and Downs syndromes, to detect genetic abnormality.
Fortune telling" in the Chinese culture: indication of the past and
future based on the patterns.
It is believed that palm print is unique to individuals. They remain
unchanged throughout at least a certain period during the adult
life of an individual.
Advantages
High distinctiveness
High permanence
High performance
Non – intrusiveness
Low resolution imaging
User – friendly
Low price palmprint devices and low setup cost
Highly stable
Palmprint Features
Principal lines: the heart line, the life line and the head line.
Regions: finger-root (I), inside region (II) and outside region (III)
Datum points: end-points across the palm and their mid-point.
Other features:
Geometry features: width of the palm, length of the palm and the area of
the palm.
Wrinkle features: these are lines other than the principal lines. They tend to
be thinner and more irregular. They are classifieds as coarse wrinkles and
fine wrinkles.
Delta point features: these are defined as the center of a delta-like region
in the palm print.
Minutiae features: similar to finger print type of features.
Palmprint
Palmprint Features
Palmprint Acquisition: Offline
Offline
Inked palm images
Scanned into 400x400, 256 gray levels
at 100 or 200 dpi (low resolution)
Area of the palm (Region of interest) is
“cropped” manually.
Palmprint Acquisition: Online
Block Diagram
Pre-processing: ROI Detection
The following steps are followed to extract the central part of the palmprint image
as ROI.
Step 1: Convert the palm image to a binary image. Gaussian smoothing is used to
enhance the image.
Step 2: Apply boundary-tracking algorithm to obtain the boundaries of the gaps
between the fingers. Since the ring and the middle fingers are not useful for
processing. Therefore, boundary of the gap between these two fingers is not
extracted.
Step 3: Determine palmprint coordinate system by computing the tangent of the
two gaps with any two points on these gaps. The Y-axis is considered as the line
which joining these two points. To determine the origin of the coordinate system,
midpoint of these two points are taken through which a line is passing and the
line is perpendicular to the Y-axis.
Step 4: Finally, extract ROI for feature extraction which is the central part of the
palmprint.
Pre-processing: ROI Detection
Feature Representation
A palmprint can be represented by some line features from a low-
resolution image.
Algorithms such as the stack filter are able to extract the principal lines.
However, these principal lines are not sufficient to represent the
uniqueness of each individual’s palmprint because different people
may have similar principal lines in their palmprints.
Fig. 1 demonstrates this problem by showing nine different palmprint
samples that have similar principal lines. In addition, some palmprint
images do not have clear wrinkles (see Fig. 2).
As a result, we try to extract texture features from low-resolution
palmprint images, and we propose a 2D Gabor phase coding scheme
for palmprint representation, which has been used for iris recognition.
Feature Representation
The circular Gabor filter is an effective tool for texture analysis and has
the following general form:
• To make it more robust against brightness, a discrete Gabor filter, , is
turned to zero DC (direct current) with the application of the following formula:
Feature Representation
Matching
Given two data sets, a matching algorithm determines the degree of similarity
between them.
To describe the matching process clearly, we use a feature vector to
represent image data that consists of two feature matrices, real and
imaginary.
A normalized Hamming distance is adopted to determine the similarity
measurement for palmprint matching. Let P and Q be two palmprint feature
vectors.
The normalized hamming distance can be described as
