Biometric key generation using crow algorithm

Indonesian Journal of Electrical Engineering and Computer Science
Vol. 21, No. 1, January 2021, pp. 208~214
ISSN: 2502-4752, DOI: 10.11591/ijeecs.v21.i1.pp208-214  208
Journal homepage: http://ijeecs.iaescore.com
Biometric key generation using crow algorithm
Zied O. Ahmed, Abbas Akram Khorsheed
Computer Science Department, Mustansiriyah University, Baghdad, Iraq
Article Info ABSTRACT
Article history:
Received May 7, 2020
Revised Jul 6, 2020
Accepted Jul 27, 2020
The researchers have been exploring methods to use biometric characteristics
of the user as a replacement for using unforgettable pass-word, in an attempt
to build robust cryptographic keys, because, human users detect difficulties
to call up long cryptographic keys. Biometric recognition provides an
authentic solution to the authentication of the user problem in the identity
administration systems. With the extensive utilization of biometric methods
in different applications, there is growing concern about the confidentiality
and security of the biometric technologies. This paper proposes biometric
based key recreation scheme. Since human ears are not correlated. Until now,
the encryption keys are generated using a swarm intelligence approach.
Collective intelligence of simple groups of autonomous agents have been
emerged by swarm intelligence. The crow search algorithm which is known
as (CSA) is a new meta-intuitive method assembled by the intelligent group
behavior of crows. Despite that CSA demonstrates important features, its
search approach poses excessive challenges while faced with great
multimodal formularization.
Keywords:
Biometric
Crow search algorithm
This is an open access article under the CC BY-SA license.
Corresponding Author:
Zied O. Ahmed
Computer Science Department
Mustansiriyah University, Baghdad, Iraq
Email: zied_othman@uomustansiriyah.edu.iq
1. INTRODUCTION
Biometrics can be defined as the science of creating an individual's identification based on a
person's physical, chemical or behavioral features [1]. Due to the properties of the biometric particular nature
[2] and the undeniable it offers [3], biometry can be often used to improve the general security of the system
to which it is applied: authentication system encryption system of biometric. The authentication of biometric
can be explained as the procedure of verifying the singularity of human in accordance with his/her
physiological features or behavioral characteristics [4]. Fingerprint, an iris or a face are the Physiological
characteristics, which refers to something that represent an individual. Whereas speech, keystroke and
signature dynamics are the communicative behavioral characteristics which refers to something that can be
done by individual. As stated by Biggio [5], the common integrated biometric authentication system works as
the following procedure. An individual provides his/her identity to access to any resource. The sample of the
user’s biometric has been taken by the sensor. Properties are obtained by the sample and a resemblance
percentage can be calculated between the two samples, first sample is the biometric which has been provided
and the second is the sample that has been stored in the biometric pattern data-base correspondent to the
identity of the user which has been provided. The resemblance percentage is contrasted to the threshold in
order to identify the user is as real or a fake. According to this determination, the access is granted or rejected
to the resources. Biometric encryption systems including the key generation and the key binding systems,
incorporate the highest level of security provided by rejection supplied by biometry and cryptography. The
systems of key generation can be known as those systems that create a stable encryption key obtained by

Indonesian J Elec Eng & Comp Sci ISSN: 2502-4752 
Biometric key generation using crow algorithm (Zied O. Ahmed)
209
biometric data [6, 7]. The systems of key binding systems can be known as the systems that can bind an
encryption key, which has been generated randomly to the biometric pattern [8, 9]; the bound key is left to
apply upon a logical presentation of the proper biometric model. The biometric samples that have been stored
create a risk to the users’ privacy.
The crow search algorithm (CSA) can be defined as a meta-heuristic way where users imitate the
behavior of being intelligent in a pack of crows. The results which have been published show the capability
to find a solution for a different complicated engineering optimization difficulties. Some cases comprise
image processing in addition to its application to water resources. Although it has important results, its search
scheme poses considerable challenges when faced with the highest multi-modal formulations [10]. Chang et
al. [11] have declared their proposal about a framework for generating a stable cryptographic key based on
single mode biometric properties which are volatile in nature. The main article of their research is the
procedure to producing properties of distinguishable biometric, which results in a strong cryptographic key.
Although the framework performance which has been proposed can be estimated with a face data-base that
including the expressions of the face and the head movement differences, the authors have announced that it
applies to other biometric methods as well as the framework.
According to the Reed-Solomon algorithm, Wu et al. [12] have improved a new system for face
biometric encryption by using 128 dimensional primary component investigation vector and error correction
codes which has known as (ECC). In the decryption step, a biometric key is resulted by using the look-up
table generated during the step of encryption, while the final key can be achieved by utilizing the biometric
key with the ECC. Feng Hao et al. [13] have submitted a cryptographic key method based on biometric by
using the iris characteristic. According to the authorized codes of iris, a periodic string of binary entitled as a
biometric key has been generated in order to be more authentic. Extra data of error correction that does not
unvei006C the key and can be accumulated in a type of tamper-resistant because a smart card is utilized to
generate the key according to an individual's iris image. Beng.A et al. [14] have proposed a randomized
biometric assistant according to a biometric key generation pattern. This technique comprises a password
backup structure and a random feature discretization procedure. First, it allowed errors to be minimized, any
other way, the later tested intra-class differences of biometric data compared with the minimum. The
unsystematic biometric assistant demonstrated which it is easy to override a biometric key as soon as the key
becomes clear.
Mishra and Bali [15], have presented an application of genetic Algorithm in the cryptography field.
Key selection is a selection procedure in which keys has been classified according to their suitability which
has happened within the public key encryption. Ultimately, it generated the most random and unrepeatable
final keys, which have raised the robustness and security of the keys. Jhajharia et al. [16] have proposed an
algorithm of public key cryptography (PKC) based on the composite principle of the two developed
algorithms, which are particle swarm optimization (PSO) and genetic algorithm (GA) consecutively. These
algorithms can be used to generate the more relevant fine fit keys in the field including the finest keys with
the highest strength. Abu-Mouti and Elhawary [17] has declared an overview of the utilize of the literature of
an Artificial which is known as bee colony (ABC) algorithm, which is a meta-heuristic based on a population
improvement procedure obtained by the intelligent feeding behavior of honey bee flocks. The performance
features and basic properties of the ABC algorithm are also explained.
Sanaul Hoque et al. [18] have presented the direct generation of the biometric keys from live
biometrics, under certain conditions, by partitioning feature space into subspaces and partitioning these into
cells, where each cell subspace contributes to the overall key generated. They assessed the presented
technique on real biometric data, instead of both genuine samples and attempted imitations. Experimental
results have proved the reliability in possible practical scenarios for this technique. A et al. [19]. The
technique engrosses a randomized feature discretization process and a code redundancy construction. The
former method controls the intra-class variations of biometric data to the nominal level and the latter reduced
the errors even more. The randomized biometric feature was proved as a simple technique, when the key was
conciliated The projected technique was assessed in the context of face data based on a subset of the facial
recognition technology (FERET) database.
2. THE CROW SEARCH ALGORITHM (CSA)
The pensive observation of some living beings’ behavior can demonstrate how they plan their
ordinary behavior to algorithmic standards. These algorithms are comprehensive optimization meta-heuristics
which can be basically composed by selecting the most usefully scheme and a random structure. Previous
guidelines eliminate both loss of diversity and algorithm to limit both local loss and the algorithm that
combines optimal use. Global optimality success can be obtained by a good stability of the use and
investigation [20]. The crow search algorithm (CSA) is a recent metaheuristic method based on the intelligent

 ISSN: 2502-4752
Indonesian J Elec Eng & Comp Sci, Vol. 21, No. 1, January 2021 : 208 - 214
210
group behavior of crows. Although CSA presents interesting characteristics, its search strategy presents great
difficulties when it faces high multi-modal formulations [21]. The recent metaheuristic algorithm which is
crow search algorithm has been developed by Alireza Askarzadeh [22], it is inspired by the crow intelligent
behavior. Normally, crows prove the intelligence behaviors as self-realization, discriminating faces,
stimulating the influx of possibly hostile ones, advanced ways of communication, and reminding the secret
place of food after a period of time. All these behaviors can be connected to the truth that the crows’ brain-to-
body ratio is less than that of the human brain which can be considered one of nature's smartest birds a little
bit [23].
The CSA evolutionary procedure imitates the crows’ behavior of hiding and regaining more food.
As an algorithm that based on population, the flock’s size is confirmed by N individual (crow) whose
problem size is n-dimensional n. The position k of Crow i in any given iteration k is described in (1) and
represent a possibly problem solution [23].
𝑋𝑖,𝑘 = [𝑥𝑖,𝑘
1
,𝑥𝑖,𝑘
2
, … , 𝑥𝑖,𝑘
𝑛
] 𝑖 = 1,2,3 . . , 𝑁; 𝑘 = 1,2, … 𝑚𝑎𝑥 𝐼𝑡𝑒𝑟 (1)
where maxIter is the utmost iteration within the procedure. Each crow individually is supposed to be able to
remember the most suitable visited the location of Mi,k to conceal food until the present iteration which
described in (2) [23].
𝑀𝑖,𝑘 = [𝑚𝑖,𝑘
1
,𝑚𝑖,𝑘
2
,… , 𝑚𝑖,𝑘
𝑛
] (2)
The location of each iteration is moderated based on Pursuit and Evasion behaviors[24]. Pursuit: A
crow j comes after Crow i to find its secret place. While crow i does not watch out the existence of the
following crow, as a result of crow j to obtain the purpose [24]. Evasion: The crow i notices crow j existence
and the crow deliberately take a random orbit to protect its food. This behavior is simulated by applying a
random motion in the CSA [24].
The concepts of crow search algorithm [25]:
a) Crows can live in the pattern of flocks.
b) Crows remember the secret places location.
c) Crows often follow each other to steal.
d) Crows preserve their hiding places from a prospect thievery
Figure 1 resent pseudo code for the algorithm of crow search.
Figure 1. The algorithm of crow search pseudo code [25]

211
3. THE PROPOSED SYSTEM
In this system more than one level is applied to generate the key that is used for encryption, the key
is generated using crow which will use the unique features of the person to extract features that is used to
generate key. Figure 2 shows the proposed system architecture and algorithm shows (1) the main phases in
the proposed key generation algorithm.
Figure 2. Proposed system architecture
Algorithm 2 proposed key generation algorithm
Input: ear image
Output: unique features
Start
Step1: read the input image
Step2: apply pre-processing phase which is consist of three internal steps:
e) RGB to gray conversion Using the following equivalence:
L=0.299R+0.587G+0.114B
f) Blurring using Gaussian filter
g) Thresholding
Step3: detect the region of interest using contour algorithm.
Step4: extract the features using LDA.
Step5: generate the encryption key using meerkat swarm algorithm.
End
This system consists of five main phases:
3.1. Pre-processing phase
This phase consists of three internal steps which used to prepare the data set for further processing
each step in this phase applied for a specific task:

 ISSN: 2502-4752
212
a) RGB to gray conversion: in this step the entered colored image will be transformed into gray scale to
decrease the amount of information processed in the system and remove noise.
b) Blurring: this is done via applying the 2D Gaussian filter to enhance the image and reduce the noise
within that image.
c) Thresholding: which will convert the image to binary which will be entered to the region of interest
extraction using contour which need binary inputs since it senses the black spots.
3.2. Region of interest (ROI) detection phase
The detect part need to be bounded since it is the only part that will need to extracted using feature
extraction algorithm, the detection of region of interests done using contour algorithm which will bounded
the ear or eye and discard other information on the image.
3.3. Feature extraction phase
This phase aims to extract features from bounded region of interest founded by contour this is done
via algorithm LDA.
3.4. Key generation phase
For the ear image 7 keys is generated for each person. To generate these keys, we use crow search
algorithm (CSA). In the beginning CSA generate initial random crow positions of keys from features
extracted from ears, the generation process done by using logistic map function. The logistic map is
evaluated by:
xn+1 = r xn (1-xn) (3)
where xn is a number from zero to one representing the current population to the maximum possible
population ratio. After generate initial population algorithm compute fitness for population to evaluate the
position of crow using fitness function, the strength value can be calculated for each iteration individually.
This value is calculated according to of the maximum repeated symbol. The suitability function can be
described as the following:
F= n + (€ / m) (4)
Where
F stands for Fitness Function.
n stands for the entire number of symbols used in key formation.
m stands for the maximum appeared symbol percentage.
€ stands for the ideal percentage for each symbol.
For each crow generate random value ri, compare ri with awareness probability (AP). If ri greater
than AP then calculate new position for crow randomly, else calculate new position for crow via levy-flight
function. After that chick feasibility of new position, and calculate fitness for new positions. These steps
repeated until reach max iteration. After generate best key from extracted feature the proposed system tests
this key.
3.5. Test generation keys
The proposed system test kit, produced by CSA (arbitrarily long) can be described as a statistical
package composing of seven test developed to test whether a random binary sequence. These tests focus on
various non-random types that may be found in a sequence. Several tests can be divided into various sub-
tests. These 7 tests are as the following:
a) The Test of Monobit (the frequency)
b) Test of Frequency through a Block.
c) Test of Runs
d) Test of Binary Matrix Rank
e) Test of Discrete Fourier Transform (Spectral)
f) Test of Non-Overlapping Template Matching
g) Test Overlapping Template Matching

213
4. RESULT DISCUSSION
The proposed system was applied to 46 people, each person has 7 snapshots for ear, each snapshot
offers a key to the person. Table 1 shows the result of test keys which are 7 keys for one person, frequency
test results in columns, test of frequency within a block results, run test results, binary matrix sequence test
results, discrete fourier transform (DFT) test results, non-overlapping template matching test results and
matching template matching test results. Two value tests for each test available test result and erfc present
error rate function results available. As shown in the Table 1, all values of erfc column in each test is greater
than 0.01, that mean the created keys are unartistic.
Table 1. Keys test results
FrequencyTest FrequencyBl
ock
RunTest RankTest DFT NonOverling OverlappingTest
Test
s
Erfc Test
s
erfc test
s
erfc tests erfc Tests erfc Tests erfc tests Erfc
0.75
0.5303
3
37.5 18.75
13
8
29.015
87
5.3964
98
14.853
7
-
44.44
87
31.429
96
0.1395
19
0.0697
6
91.159
76
45.579
88
0.87
5
0.6187
18
25.7
5
12.87
5
13
1
9.5394
23
5.3964
98
14.853
7
-
45.25
98
32.003
49
0.1715
4
0.0857
7
87.434
11
43.717
05
0.5
0.3535
53
28 14
12
6
5.2981
22
5.3964
98
14.853
7
-
45.66
53
32.290
26
0.1418
06
0.0709
03
89.068
29
44.534
15
0.37
5
0.2651
65
30.2
5
15.12
5
14
1
36.948
12
1.6797
39
2.3160
64
-
42.42
09
29.996
11
0.1440
94
0.0720
47
90.960
91
45.480
45
1.62
5
1.1490
49
45.7
5
22.87
5
13
0
9.2943
92
1.6797
39
2.3160
64
-
46.47
64
32.863
8
0.0937
75
0.0468
88
85.999
2
42.999
6
0.37
5
0.2651
65
30.2
5
15.12
5
13
4
17.160
01
1.6797
39
2.3160
64
-
46.47
64
32.863
8
0.1555
3
0.0777
65
89.239
34
44.619
67
0.5 0.3535
53 32 16
13
9
31.435
52
2.0370
79
2.7691
48
-
46.47
64
32.863
8
0.1029
24
0.0514
62
92.629
08
46.314
54
5. CONCLUSION
The work has been carried out for hundreds of samples. Each population varies greatly from another.
Key length for which test is carried out is 256 bit long. Longer key sequence will also work but time
constraint does not permit to check. Encryption and decryption are also performed. This paper proposes the
Crow Search Algorithm, and shows its potential to generate Keys from biometric features and obtain the
random keys with a smaller number of iterations.
REFERENCES
[1] A. K. Jain, A., Ross, “Introduction to Biometrics. In Handbook of Biometrics,” (Eds), Springer, 2008
[2] Y. C. Feng, P. C. Yuen, A. K. Jain, “A Hybrid Approach for Face Template Protection,” In Proceedings of SPIE
Conference of Biometric Technology for Human Identification, Orlando, USA, vol. 6944, pp. 325, 2008.
[3] P. Balakumar, R. Venkatesan, “A Survey on Biometrics-based Cryptographic Key Generation Schemes,”
International Journal of Computer Science and Information Technology & Security, vol. 2, no. 1, pp. 80-85, 2012.
[4] A. K. Jain, A. Ross, S. Prabhakar, “An Introduction to Biometric Recognition,” IEEE Transactions on Circuits and
Systems for Video Technology, vol. 14, pp. 4-20, 2004.
[5] B. Biggio, “Adversarial Pattern Classification,” Doctoral dissertation, University of Cagliari, Cagliari, Italy, 2010.
[6] G. I. Davida, Y. Frankel, B. J. Matt, “On Enabling Secure Applications through Off-Line Biometric Identification,”
In Proceedings of the IEEE Symposium on Privacy and Security, pp. 148-157, 1998.
[7] Y. J. Chang, W. Zhang, T. Chen, “Biometrics-based Cryptographic Key Generation,” In Multimedia and Expo,
ICME'04. 2004 IEEE International Conference, vol. 3, pp. 2203-2206, 2004.
[8] A. Juels, M. Sudan, “A fuzzy vault scheme,” In Proc. IEEE Int. Symp. Information Theory, IEEE Press, pp. 408,
2002.
[9] Y. Dodis, L. Reyzin, A. Smith, “Fuzzy Extractors: How to Generate Strong Keys from Biometrics and Other Noisy
Data,” In Proceedings of the Eurocrypt 2004, pp. 523-540, 2004
[10] Askarzadeh, A., “Capacitor placement in distribution systems for power loss reduction and voltage improvement: a
new methodology,” IET Generation, Transmission & Distribution, vol. 10, no. 14, pp. 3631–3638, 2016.

 ISSN: 2502-4752
214
[11] Y. J. Chang, W. Zhang, T. Chen, “Biometrics-based Cryptographic Key Generation,” In Multimedia and Expo,
ICME'04. 2004 IEEE International Conference, vol. 3, pp. 2203-2206, 2004.
[12] L. Wu, X. Liu, S. Yuan, P. Xiao, “A Novel Key Generation Cryptosystem based on Face Features,” In Signal
Processing (ICSP) 2010 IEEE 10th International Conference on, pp. 1675-1678, 2010.
[13] Feng Hao, Ross Anderson and John Daugman, “Combining Crypto with Biometrics Effectively”, IEEE
Transactions on Computers, vol. 55, no. 9, pp. 1081 - 1088, 2006.
[14] Beng.A, Jin Teoh and Kar-Ann Toh, “Secure biometrickey generation with biometric helper,” in proceedings of
3rd IEEE Conference on Industrial Electronics and Applications, pp. 2145-2150, 2008.
[15] S.Mishra and S.Bali, “Public key cryptography using genetic algorithm,” International Journal of Recent
Technology and Engineering, vol. 2, no. 2, pp. 150-154, 2013.
[16] S.Jhajharia, S.Mishra and S.Bali, “Public key cryptography using particle swarm optimization and genetic
algorithm,” International Journal of Advanced Research in Computer Science and Software Engineering, vol. 3,
no. 6, pp. 832-839, 2013.
[17] F.S.A.-Mouti, M.E.Elhawary, “Overview of artificial bee colony algorithm and its applications,” IEEE
International Conference–System Conference, pp. 1-6, 2013.
[18] Sanaul Hoque, Michael Fairhurst and Gareth Howells, “Evaluating Biometric Encryption Key Generation Using
Handwritten Signatures”, in Proceedings of the 2008 Bio-inspired, Learning and Intelligent Systems for Security,
pp.17-22, 2008.
[19] Beng.A, Jin Teoh and Kar-Ann Toh, “Secure biometric-key generation with biometric helper”, in proceedings of
3rd IEEE Conference on Industrial Electronics and Applications, pp. 2145-2150, 2008.
[20] Ahmed T. Sadiq Al-Obaidi, Hasanen S. Abdullah, Zied O. Ahmed, “Meerkat Clan Algorithm: A New Swarm
Intelligence Algorithm,” Indonesian Journal of Electrical Engineering and Computer Science, vol. 10, no. 1, pp.
354-360, 2018.
[21] Primitivo Díaz, Erik Cuevas, “An Improved Crow Search Algorithm Applied to Energy Problems”, Energies, vol.
11, no. 571, 2018.
[22] Askarzadeh, A., “Capacitor placement in distribution systems for power loss reduction and voltage improvement: a
new methodology” IET Generation, Transmission & Distribution, vol. 10, no. 14, pp. 3631–3638, 2016.
[23] Emery, N. J., & Clayton, N. S., “The Mentality of Crows: Convergent Evolution of Intelligence in Corvids and
Apes,” Science, vol. 306, no. 5703, 2004.
[24] M. K. Marichelvam, K. Manivannan, M. Geetha, “Solving Single Machine Scheduling Problems using an
Improved Crow Search Algorithm,” International Journal of Engineering Technology Science and Research
IJETSR, vol. 3, no. 12, 2016.
[25] Askarzadeh, A., “A novel metaheuristic method for solving constrained engineering optimization problems: Crow
search algorithm,” Computers and Structures, vol. 169, no. 1, pp. 1-12, 2016.

Biometric key generation using crow algorithm

More Related Content

Similar to Biometric key generation using crow algorithm (20)

More from nooriasukmaningtyas (20)

Recently uploaded (20)

Biometric key generation using crow algorithm