1. PATTERN RECOGNITION TECHNIQUE
FOR PREDICTION OF BLOOD TYPE
FROM INDIVIDUALS FINGERPRINT
R.RAJKUMAR
Assistant Professor
Department of Computer Science and Engineering
SRM University, KTR
rajkumar.ra@ktr.srmuniv.ac.in
Guided by
Dr. V. GANAPATHY
Professor
Department of Information Technology
SRM University
2. ABSTRACT
Fingerprints are most reliable tool for
personal identification. The finger prints remain
unchanged in an individual throughout his/her
life time.
For the pattern recognition of fingerprints,
machine learning techniques for prediction of
blood types are proposed.
The fingerprint study explores the relation
between blood type and fingerprint of individuals
for the applications of forensic science, medical
treatments and accident fields
3. PROBLEM DESCRIPTION
Studies say that, individual’s blood type
plays a major role in her/his food habits,
behavior, memory, stress tolerance and
many other personal characteristics.
The dermatoglyphics is the study of
pattern of fine ridges on finger prints.
The type of finger print is unique and it
is based on genetic characters of each
individual. Still the relationship between
finger print and blood type is unexplored.
5. Various patterns of finger prints
Analysts use the general pattern type
(loop, whorl or arch) to make initial
comparisons and include or exclude a
known fingerprint from further analysis.
The following 11 patterns that appear on
our fingerprints generally explain one’s
personality; each pattern on each different
finger may be explained differently with
different analysis.
6. i) Simple Arch Patterns
Form: hill-shaped, curved top, no triangle was
formed in with the shape.
ii) Tented Arch Patterns
Form: like a camping tent with a sharp tip top.
iii) Ulnar Loop Patterns
Form: like a waterfall flowing towards the little
finger with triangular points.
Various patterns of finger prints
7. iv) Radial Loop Patterns
Form: The opposite of ulnar loop, the
“waterfall” flows toward the thumb.
v) Concentric Whorl Patterns
Form: Lines starting from the center of the
small circle, the lines on finger tip appear to
be a complete circle and spread out like
concentric circles. with two triangular
points.
Various patterns of finger prints
8. vi) Spiral Whorl Patterns
Form: A spiral pattern starting from the
center and move outward, has two
triangular points.
vii) Press Whorl Patterns
Form: Similar to the whorl pattern, but the
circle turns into a long oval shape, has two
triangular
Various patterns of finger prints
9. viii) Imploding whorl Patterns
Form: Tai Chi-like patterns in the middle,
surrounded by multi-layers of circle.
ix) Composite Whorl Patterns
Form: Tai Chi-lke pattern without multi-layers
or circle surrounding it.
Various patterns of finger prints
10. x) Peacock’s Eye Patterns
Form: From the center it looks like a peacock’s eyes
and lips; the center consists of more than one
circle or spiral, the end of each ring is connected
in a straight line. It has two triangular points; one
further and the other closer to the center.
xi) Variant Patterns
Form: Often has combination of two or more of
whorls, ulnar loops, or simple arches, with two.
Various patterns of finger prints
11. We have taken 10 thump finger print
samples for each of blood types
(A/B/AB/O) for a total subjects of 40
people.
We used Matlab Image Processing
toolbox for preprocessing of the images
and template matching.
Experiment
12. Template-based template matching may
potentially require sampling of a large
number of points, it is possible to reduce the
number of sampling points.
For reducing the resolution of the search
and template images by the same factor and
performing the operation on the resultant
downsized images (multi-resolution, or
Pyramid (image processing))
Template matching : Pre-processing
13. The partitioning clustering approach partitions
the database of n-objects into k-number of
predefined clusters where each partition represents
a cluster and k cluster. It also satisfies two
conditions
(i) each cluster must have at least one object,
(ii) each object must belong to exactly one cluster.
Method
15. Several categories have been defined in
the Henry system: whorl, right loop, left loop,
arch, and tented arch.
Most algorithms are using minutiae, the
specific points like ridges ending, bifurcation.
Only the position and direction of these
features are stored in the signature for further
comparison.
Algorithm
16. Code for load new image to Matlab.
I=imread(‘Fingerprint.bmp');
imshow(I)
set(gcf,'position',[1 1 600 600]);
17. Binarize the image
We binarize the image. After the operation,
ridges in the fingerprint are highlighted with
black color while furrow are white.
Code for binatize the image:
J=I(:,:,1)>160;
imshow(J)
set(gcf,'position',[1 1 600 600]);
18. Thinning
Thinning is to eliminate the redundant
pixels of ridges till the ridges are just one
pixel wide.
Code for thining the image:
K=bwmorph(~J,'thin','inf');
imshow(~K)
set(gcf,'position',[1 1 600 600]);
20. Cluster the finger prints Using K-Means
Clustering is a way to separate groups of
objects. K-means clustering treats each object
as having a location in space.
It finds partitions such that objects within
each cluster are as close to each other as
possible, and as far from objects in other
clusters as possible.
K-means clustering requires that you specify
the number of clusters to be partitioned and a
distance metric to quantify how close two objects
are to each other.
21. K- means in matlab
ab = double(lab_he(:,:,2:3));
nrows = size(ab,1);
ncols = size(ab,2);
ab = reshape(ab,nrows*ncols,2);
nColors =3;
% repeat the clustering 3 times to avoid local
minima
[cluster_idx, cluster_center] =
kmeans(ab,nColors,'distance','sqEuclidean',
'Replicates',3);
22. Results
Blood type Whorls Loops Arches
O 8 % 67 % 25 %
A 71.2 % 9.81 % 9%
B 26.5 % 3.45 % 69 %
AB 82 % 6 % 13 %
Mean of blood types
23. Results
Blood type Whorls Loops Arches
O 8 % 67 % 25 %
A 71.2 % 9.81 % 9%
B 26.5 % 3.45 % 69 %
AB 82 % 6 % 13 %
Mean of blood types
24. Results
O has 66-70 % Loops
A has 70- 75 % Whorls
B has 65- 73% of Arches
AB has 80- 85% whorls
25. Conclusion
We have used unsupervised clustering method to
pattern reorganization among finger prints of different
types of blood types.
In the result the finger print images are grouped with
their respective blood type like A, B, O and AB. Some of
finger prints are in the middle of the cluster.
Fingerprint loops are found to be associated more
with O blood type, whorls with AB blood type, arches
with B blood type and high frequency of whorls
associated with A blood type.
If we use supervised learning method with the 11
types of finger prints we can increase the chances to
predict the blood type by finger print.
26. References
• [1] Cattell R. B, Boutourline Y, Blood group and
personality traits, American Journal of human genetics,
Vol 16, No.4(December, 1984.
• [2] http://dermatoglyphics.org/fingerprint-radial-loop
• [3] Raghavan R and John Singh K, Necessity of different
patterns of fingerprint and its applications: A study,
International Journal of Applied Engineering Research
ISSN 0973-4562 Volume 11, Number 8 (2016) pp 5787-
5790.
• [4] http://www.forensicsciencesimplified.org
27. • [5] López-Escobar Saúl, Carrasco-Ochoa J.A., Martínez-
Trinidad J. Fco., “Global k-Means with Similarity
Functions," CIARP 2005, LNCS 3773, pp. 392-399,
2005.
• [6] Gongping Yang,1 Guang-Tong Zhou,1 Yilong Yin,1
and Xiukun Yang, K-Means Based Fingerprint
Segmentation with Sensor Interoperability, EURASIP
Journal on Advances in Signal Processing.Volume 2010,
Article ID 729378, 12 pages.
• [7] R. C. Dubes and A. K. Jain, Algorithms for Clustering
Data, Prentice Hall, Upper Saddle River, NJ, USA, 1988.
References