Bulletin of Applied Mathematics and Mathematics Education

Volume 3, Number 2, 2023 pp. 69-78

https://dx.doi.org/10.12928/bamme.v3i2.9949
e-ISSN 2776-1029, p-ISSN 2776-1002

Identification of fingerprint image with Minkowski distance

algorithm approach
Wisnu Catur Rachmad Tulloh, Dian Eka Wijayanti*
Universitas Ahmad Dahlan, Jl. Jend. Ahmad Yani, Tamanan, Banguntapan, Bantul, DIY 55711, Indonesia

*Corresponding E-mail: [email protected]

ARTICLE INFO ABSTRACT

In the digital era, fingerprint identification plays a critical role in

Article History information technology administration. Various studies have been
Received 14 December 2023 conducted to improve the fingerprint identification process, but
Revised 29 December 2023 there are still cases of identification failures that are fatal. This
Accepted 29 December 2023 research discusses fingerprint identification with the Minkowski
distance method. The data of fingerprint are taken from Mathematics
Keywords students and the Kaggle site. Data analysis includes the steps of
Data analysis image retrieval, dimensioning, conversion to gray scale, pattern
Fingerprint identification matching, and accuracy measurement. Results show an
Minkowski distance improvement in data accuracy with a structured approach to data
System security capture and preprocessing. Results from primary data obtained an
accuracy of 56.67% while from secondary data obtained an accuracy
How to cite this article: of 93%.
Tulloh, W. C. R., & Wijayanti, D. E.
(2023). Identification of This is an open access article under the CC–BY-SA license.
fingerprint image with
Minkowski distance algorithm
approach. Bulletin of Applied
Mathematics and Mathematics
Education, 3(2), 69-78.

Introduction
The rapid growth in technology and information systems in the digital age is undergoing significant
changes. Easy access to information through the internet has become a primary need, but the
increasing use of digital technology also brings the risk of increased threats to security. Therefore,
it is important to prioritize digital security protection that includes aspects of personal data
protection, network security, and information system security(Gani, n.d.; Mayamin & Usuluddin,
2023; Nuriadin et al., 2021; Nuriadin, 2021).
Fingerprint sensors, as part of management information systems, play a role in the
management of personal information. In the digital age, technological advancements impact on the
vulnerability of information security. In the digital financial sector, this poses a threat of loss of
assets or leakage of customer data, requiring simple and valid identification. Fingerprint sensors
have several advantages, such as simplifying the identification process and data storage, while
increasing the security of personal information(Emelda, n.d.; Rizaldi et al., n.d.; Nurliza & Widodo,
2018).
Fingerprint identification can be used to improve security and prevent data leakage. A study
conducted by Satria et. al. (2017) showed that the fingerprint system can prevent personal data

69
BAMME Vol. 3 No. 2, 2023, pp. 69-78

leakage up to 99.9%. In addition, another study conducted by Arifandi (2023) showed that a door
security system using fingerprints based on Arduino Uno ATMega328 and SMS Gateway can
identify residents of the house and provide solutions to homes left by their owners. In the study,
the test results show access to open the door using fingerprints from outside can run well and
homeowners can easily get information when someone breaks into the door by force (Arifandi,
2019; Aziz, n.d.; Gusri et al., n.d.; Hartono et al., 2023; Hermawan et al., 2023).
Various methods can be used in fingerprint identification, one of which is the Minkowski
distance. The Minkowski distance method acts as an important metric for vector spaces, serving as
a norm in the space, encompassing the generalized forms of Euclidean and Manhattan distances.
The use of Minkowski distance in the fingerprinting approach contributes to the formation of a
unique signature, ensuring a distinctive and non-copyable identification. This methodology, which
is an integral part of vector space analysis, highlights the precision and complexity essential for
robust applications, making it a key element in a collection of advanced techniques in the domain
of data science and pattern recognition(Nurliza & Widodo, 2018; Pradana, 2017; Safwandi &
Muthmainnah, n.d.).
Research that adopts the Minkowski distance method to perform distance analysis between
data points in vector space can result in a deeper understanding of the structure and distribution
of the data, open up the potential for discovering relevant patterns, and improve the accuracy of the
analysis in the context in question. The Minkowski distance method is a very suitable method for
translating distance analysis between data points, because it provides flexibility and generalization
in measuring distance. The steps of this research involve calculating distance using the Minkowski
distance formula to describe the relationships and patterns that exist in the data(Khairunnisa et al.,
n.d.; Nishom, 2019; Safwandi & Muthmainnah, n.d.; Thant & Aye, 2020).

Method
This research adopting the Minkowski distance method to perform distance analysis between data
points in vector space can lead to a deeper understanding of the structure and distribution of the
data, open up the potential for discovering relevant patterns, and improve the accuracy of the
analysis in the context in question. The Minkowski distance method is a very suitable method for
translating distance analysis between data points, as it provides flexibility and generality in
measuring distance. The steps of this research involved calculating distances using the Minkowski
distance formula to illustrate the relationships and patterns present in the data.

Data collection methods

Experiment research
In the data collection process, primary data was used by randomly collecting 30 fingerprints from
students of the mathematics study program at Ahmad Dahlan University. In addition, an additional
6 reference data were also obtained to be included in the calculation of the Minkowski distance
method. This method requires reference data as a reference in measuring the distance between
primary data. Thus, the fingerprint data collected from these students will be analyzed using the
Minkowski distance method by considering the 6 references data that have been previously
determined. This aims to enable proper comparison between the primary data and the reference
data, thus facilitating an accurate analysis process in the context of the Minkowski distance method.
Literature study
The data used in this research is secondary data obtained from the Kaggle platform. The data

70 10.12928/bamme.v3i2.9949
 e-ISSN 2776-1029 p-ISSN 2776-1002

retrieval process was carried out with respect to the distance in which the photos were taken and
involved cropping the data. This approach allows for a more structured selection and customization
of data from such secondary sources to support proper analysis in a pre-defined method.

Techniques for analyzing data

Data analysis is an essential component of the research framework. The reliability of resolving the
issues in focus and the conclusions that result from a study have a substantial dependence on the
integrity and completeness of the data analysis conducted. The analysis methods and procedures
applied in this study included the following steps.
Fingerprint image capture
At this stage, 36 images are taken, consisting of 6 fingerprint reference templates (guide1.jpg -
guide6.jpg) and 30 input templates to be tested (finger.1.jpg - finger.30.jpg). Subsequently, these
images are stored for further processing.
The process of customizing the image dimensions to 4 × 4 pixels
This step is designed to normalize the input image to be analyzed.
Image conversion from RGB color mode to grayscale
This is necessary so that the image can be processed using the chosen method.
Pattern matching
In this phase, processing is performed between the reference sample template and the input
template that has been taken. Each template, be it the reference image or the cropped input image,
is analyzed comparatively one by one. To assess the similarity between the two templates, the
Minkowski distance method is used. This method calculates the distance value between the two
templates being compared. An input template will be classified as similar to the reference template
if the distance value between the two is equal to, or less than 30.
Accuracy measurement
In this step, the accuracy of the Minkowski distance method will be evaluated to determine whether
or not this method is suitable in the template matching process (See Figure 1).

Minkowski distance method and accuracy technique

The Minkowski Distance method has a formula that integrates the characteristics of Euclidean
Distance and Manhattan Distance with p as a parameter. When the value of 𝑝 = 1, it has the same
formula as Manhattan Distance, while if 𝑝 = 2, the formula tends to be Euclidean Distance. The
general formula of Minkowski Distance is,
%
# "
𝑑(𝑥, 𝑦) = ((|𝑥! − 𝑦! |" )
!$%
(1)

The calculation process to evaluate the accuracy of a method or procedure that has been
implemented (See Figure 1). This accuracy analysis aims to measure the extent to which the results
obtained are in accordance with the expected standards or meet predetermined criteria. This stage
makes it possible to identify the advantages and disadvantages of a method or technique used in
the context of data processing or analysis being carried out. The general formula for accuracy is,
∑𝑚𝑎𝑛𝑦 𝑠𝑢𝑐𝑐𝑒𝑠𝑠𝑒𝑠,
𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦(%) = × 100%
∑𝑡𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑑𝑎𝑡𝑎

Identification of fingerprint image (Tulloh & Wijayanti) 71

BAMME Vol. 3 No. 2, 2023, pp. 69-78

(2)

Figure 1. Minkowski distance method flowchart

Results and discussion

The data that has been taken is then converted into a matrix in the RGB color scheme, then
converted into a grayscale matrix. Furthermore, from the initial matrix that has a size of 𝑚 × 𝑚,
dimensional adjustments are made to a 4 × 4 matrix (See Figure 2). This step aims to minimize the
dimensions of the data so that it can be processed more efficiently and in accordance with the
predetermined analysis needs.
Matrix P1 refers to the matrix derived from guideline 1. Using the established definition of 1,
the Minkowski distance will be calculated for the matrix. This process is part of the evaluation of
the similarity or difference between the reference matrix and other matrices within the framework
of the predefined method.

72 10.12928/bamme.v3i2.9949
 e-ISSN 2776-1029 p-ISSN 2776-1002

177 175 181 181

⎡177 ⋯
⎢ 176 177 177⎤⎥
⎢ ⋮ ⋱ ⋮ ⎥
⎢171 171
⋯
175 175⎥
⎣171 170 175 175⎦

Fingerprint image An 𝑚 × 𝑚 matrix

Image. 2 Where m is the pixel size of the
image

177 133 149 184

𝑃1 = <157 107 117 174@
155 109 116 169
170 153 148 170

4 × 4 resize matrix

Figure 2. Processing fingerprint data into 4 × 4 matrix

Manual calculation as follows. Comparison of respondent 1 with guideline 1 results:

177 144 132 172

𝑅1 = ?148 91 92 144 E
127 83 91 140
146 114 129 161

177 133 149 184

𝑃1 = G157 107 117 174I
155 109 116 169
170 153 148 170
𝑝=3
"
𝑑(𝑥, 𝑦) = (|177 − 177|! + |133 − 144|! + |149 − 132|! + ⋯ + |170 − 161|! )!
𝑑(𝑥, 𝑦) = 59.9717459564981

With a similar approach, calculations were made using Guideline 1 to Guideline 6 against the
responses from Respondent 1 to Respondent 30. This step was carried out to analyze and evaluate
the suitability and similarity between each guideline and the respective responses generated by the
respondents involved in this study and the following Table 1 was obtained.

Identification of fingerprint image (Tulloh & Wijayanti) 73

BAMME Vol. 3 No. 2, 2023, pp. 69-78

Table 1. Calculation on data from respondents based on Guideline 1 to 6

Number Name Guidelines 1 Guidelines 2 Guidelines 3 Guidelines 4 Guidelines 5 Guidelines 6 Classification Results Success
1 Respondent1 59,9717 68,1699 37,9984 82,7347 50,2466 61,2808 Does not belong to guideline data works
2 Respondent2 80,7384 89,8481 78,3769 90,3676 69,093 63,3747 Does not belong to guideline data works
3 Respondent3 65,6447 71,2493 91,7409 46,9907 63,0258 42,7467 Does not belong to guideline data works
4 Respondent4 71,4142 69,0025 76,1025 64,2724 60,185 54,4752 Does not belong to guideline data works
5 Respondent5 96,3032 92,7073 128,6952 65,8542 104,6089 93,5718 Does not belong to guideline data works
6 Respondent6 37,9037 32,4189 58,3379 36,4415 53,6541 50,5340 Does not belong to guideline data works
7 Respondent7 50,2726 38,0567 56,3271 44,8584 46,5402 46,7214 Does not belong to guideline data works
8 Respondent8 34,2930 46,8687 56,3271 55,9647 35,2389 35,4968 Does not belong to guideline data works
9 Respondent9 43,4746 56,2209 34,6287 73,2616 45,2215 54,5415 Does not belong to guideline data works
10 Respondent10 31,7185 33,2918 59,3669 23,3059 34,8326 32,6972 Belonging to the guideline data Failed
11 Respondent11 23,4329 29,3833 30,1338 49,452 36,4162 45,4339 Belonging to the guideline data Failed
12 Respondent12 35,6585 40,1306 46,5394 38,8048 23,3915 26,167 Belonging to the guideline data Failed
13 Respondent13 101,2387 109,2219 98,0335 113,9386 23,3915 26,167 Belonging to the guideline data Failed
14 Respondent14 33,0653 42,1691 30,5394 57,085 31,9056 40,4521 Does not belong to guideline data works
15 Respondent15 42,2666 41,9367 68,8864 14,8639 40,2352 32,2301 Belonging to the guideline data Failed
16 Respondent16 45,1987 51,9454 47,6481 55,3524 29,1376 24,744 Does not belong to guideline data Failed
17 Respondent17 43,6056 45,9337 36,3201 51,611 22,3551 29,9599 Belonging to the guideline data Failed
18 Respondent18 39,5631 48,1102 56,5746 44,2941 37,4866 32,3082 Does not belong to guideline data works
19 Respondent19 29,9803 48,1102 56,5746 44,2941 37,4866 32,3082 Belonging to the guideline data Failed
20 Respondent20 37,7057 34,6576 50,2614 35,9368 20,8415 22,1171 Belonging to the guideline data Failed
21 Respondent21 57,8054 58,2673 85,166 24,6242 52,5613 39,5156 Belonging to the guideline data Failed
22 Respondent22 33,0214 38,0454 42,7767 42,3725 25,4195 33,7773 Belonging to the guideline data Failed
23 Respondent23 41,4974 39,9628 66,8592 15,9634 38,21804 27,63034 Belonging to the guideline data Failed
24 Respondent24 51,0921 51,1571 79,7653 23,55 47,6627 33,2665 Belonging to the guideline data Failed
25 Respondent25 46,7772 45,2199 53,8587 43,587 29,7007 25,993 Belonging to the guideline data works
26 Respondent26 46,7772 45,2199 53,8587 43,5870 29,7007 23,3426 Belonging to the guideline data works
27 Respondent27 45,9274 50,9755 41,5662 59,5204 27,3462 30,4041 Belonging to the guideline data works
28 Respondent28 38,5918 42,7663 66,9229 20,397 37,2382 22,8509 Belonging to the guideline data works
29 Respondent29 37,8877 26,0487 61,2362 21,6237 42,0035 39,996 Belonging to the guideline data works
30 Respondent30 40,7947 43,4481 65,8272 26,0928 35,2201 22,6753 Belonging to the guideline data works

From the previously tested data, the accuracy and success of the tested image matching can
be evaluated using the following methods.
∑𝑚𝑎𝑛𝑦 𝑠𝑢𝑐𝑐𝑒𝑠𝑠𝑒𝑠,
𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦(%) = × 100%
∑𝑡𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑑𝑎𝑡𝑎

17
𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦(%) = × 100%
30
𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦(%) = 56,67%

Then, we continue to analyze the following steps (See Figure 3). Matrix P1 refers to the matrix
derived from guideline 1. Using the established definition of 1, the Minkowski distance will be
calculated for the matrix. This process is part of the evaluation of the similarity or difference
between the reference matrix and other matrices within the framework of the predefined method.
The manual calculation is as follows. Comparison of respondent1 with guideline 1.

223 197 202 189

𝑅1 = <164 114 109 122@
140 110 116 115
141 109 116 124
230 210 210 191
𝑃1 = E215 119 107 170F
194 119 112 170
160 94 108 149
𝑝=3
"
𝑑(𝑥, 𝑦) = (|230 − 223|! + |210 − 197|! + |210 − 202|! + ⋯ + |149 − 149|! )!

74 10.12928/bamme.v3i2.9949
 e-ISSN 2776-1029 p-ISSN 2776-1002

𝑑(𝑥, 𝑦) = 84.22309366304167

22 255 0 239
⎡1 ⋯
⎢ 199 0 220⎤⎥
⎢ ⋮ ⋱ ⋮ ⎥
⎢ 0 0
⋯
0 0 ⎥
⎣ 0 0 0 0 ⎦

Fingerprint image An 𝑚 × 𝑚 matrix

Where m is the pixel size of the
image

230 210 210 191

𝑃1 = E215 119 107 170F
194 119 112 170
160 94 108 149

4 × 4 resize matrix

Figure 3. Further processing fingerprint data into 4 × 4 matrix

Table 2. Calculation on data from respondents based on Guideline 1 to 6

Number Name Guidelines 1 Guidelines 2 Guidelines 3 Guidelines 4 Guidelines 5 Guidelines 6 Classification Results Success
1 Respondent1 84,22 73,57 88,12 67,43 78,46 98,34 Does not belong to guideline data works
2 Respondent2 136,44 125,84 126,2 57,19 87,45 126,21 Does not belong to guideline data works
3 Respondent3 56,96 56,24 45,5 118,57 95,86 109,39 Does not belong to guideline data works
4 Respondent4 112,55 100,48 110,05 61,38 63,97 54,96 Does not belong to guideline data works
5 Respondent5 82,75 65,96 56,08 89,55 71,95 104,89 Does not belong to guideline data works
6 Respondent6 116,84 109,9 126,32 58,41 86,31 87,93 Does not belong to guideline data works
7 Respondent7 104,3 101,75 87,09 89,55 103,86 128,97 Does not belong to guideline data works
8 Respondent8 110,49 97,07 104,28 40,29 61,62 72,4 Does not belong to guideline data works
9 Respondent9 112,23 120,2 104,83 196,64 176,45 175,44 Does not belong to guideline data Failed
10 Respondent10 77,23 60,22 53,01 122,95 88,68 117,33 Does not belong to guideline data works
11 Respondent11 117,87 109,58 97,66 84,73 94,93 137,41 Does not belong to guideline data works
12 Respondent12 156,49 149,11 164,64 73,13 114,06 132,93 Does not belong to guideline data works
13 Respondent13 87,7 93,45 82,32 70,12 80,18 83,35 Does not belong to guideline data works
14 Respondent14 101,07 95,41 94,14 71,78 82,34 63,91 Does not belong to guideline data works
15 Respondent15 150,15 140,17 154,36 58,43 99,89 100,22 Does not belong to guideline data works
16 Respondent16 118,51 112,3 120,47 144,86 130,75 104,91 Does not belong to guideline data works
17 Respondent17 92,22 91,22 88,97 68,91 70,87 99,97 Does not belong to guideline data works
18 Respondent18 84,01 86,58 81,35 71,84 89,03 108,71 Does not belong to guideline data works
19 Respondent19 87,14 62,28 71,4 94,48 22,25 68,25 Belonging to the guideline data works
20 Respondent20 65,68 47,94 68,85 90,81 48,82 83,8 Does not belong to guideline data works
21 Respondent21 112,01 94,76 110,85 79,91 58,23 55,23 Does not belong to guideline data works
22 Respondent22 115,32 107,19 111,48 38,59 79,1 82,58 Does not belong to guideline data works
23 Respondent23 154,11 140,82 137,74 75,4 108,51 156,37 Does not belong to guideline data works
24 Respondent24 104,02 98,42 91,38 75,17 91,59 131,26 Does not belong to guideline data works
25 Respondent25 137,45 124,63 128,49 39,47 95,98 130,64 Does not belong to guideline data works
26 Respondent26 71,03 71,34 81,29 70,24 69,37 77,69 Does not belong to guideline data works
27 Respondent27 120,82 101,72 120,31 58,49 54,13 56,14 Does not belong to guideline data works
28 Respondent28 118,48 100,99 115,37 69,15 58,04 46,33 Does not belong to guideline data works
29 Respondent29 99,31 100,26 89,67 104,67 120,56 126,62 Does not belong to guideline data Failed
30 Respondent30 61,51 66,25 56,47 88,02 70,92 75,87 Does not belong to guideline data works

Identification of fingerprint image (Tulloh & Wijayanti) 75

BAMME Vol. 3 No. 2, 2023, pp. 69-78

With a similar approach, calculations were made using Guideline 1 to Guideline 6 against the
responses from Respondent 1 to Respondent 30. This step was carried out to analyze and evaluate
the suitability and similarity between each guideline and the respective responses generated by the
respondents involved in this study and Table 2 was obtained.
From the previously tested data, the accuracy and success of the tested image matching can
be evaluated using the following methods:

∑𝑚𝑎𝑛𝑦 𝑠𝑢𝑐𝑐𝑒𝑠𝑠𝑒𝑠,
𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦(%) = × 100%
∑𝑡𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑑𝑎𝑡𝑎

28
𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦(%) = × 100%
30
𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦(%) = 93%

From the results of the first data processing, an accuracy rate of 56.67% was obtained. While
in the second data processing, the accuracy rate increased to 93%. This significant difference is due
to the different approaches in data collection. In the first data collection, there was no consideration
of the data collection distance or cropping process on the data set. In contrast, in the second data
collection, a more structured approach was used with data collection at a uniform distance and the
application of a cropping process, which resulted in a significant increase in data accuracy.

Conclusion
The results of using the Minkowski Distance method on the first data showed an accuracy rate of
56.67%, which is significant but worth noting for further improvement. The use of the second data
that considered the capture distance and applied cropping resulted in an increase in accuracy to
93%, confirming the importance of a structured approach to data capture. A suggestion for future
research is to consider other factors that affect accuracy, such as variations in image capture angle
or more complex preprocessing techniques. Further evaluation of these factors can help improve
the accuracy of the method in fingerprint image recognition.
Suggestions for further research include several things that can be optimized. Firstly, it is
recommended to consider the use of additional data to improve a greater level of accuracy in
fingerprint recognition. Second, it is recommended to consider an approach to minimize the
fingerprint image in certain parts of the diagram to obtain a more optimized matrix. Thirdly, it is
recommended to further normalize the matrix to produce smaller values, simplifying the
calculation process. Finally, it is important to pay attention to the pixel size when photographing
fingerprint data, considering it as a key factor in image capture for more accurate analysis. With
these suggestions in mind, future research is expected to make significant progress in the
development of more reliable and efficient fingerprint recognition methods.

References
Arifandi, M. (2019). Prototype Sistem Keamanan Pintu Menggunakan Sidik Jari (Fingerprint)
Berbasis Arduino Uno ATMega328 dan SMS Gateway.
https://www.researchgate.net/publication/340165694
Aziz, N. (n.d.). Pemanfaatan teknologi internet dalam pendidikan.
Emelda. (n.d.). Pengaruh Absensi Fingerprint dan Sanksi Hukuman Terhadap Disiplin
Kerja Pegawai Pada Dinas Perdagangan Provinsi Sumatera Selatan. STIE Rahmaniyah Sekayu.
Gani, A. G. (n.d.). Pengenalan teknologi internet serta dampaknya.

76 10.12928/bamme.v3i2.9949
 e-ISSN 2776-1029 p-ISSN 2776-1002

Gusri, M., Dan, A., & Kadarisman, Y. (n.d.). Penggunaan Teknologi Internet Dikalangan Mahasiswa
Fakultas Ilmu Sosial Dan Ilmu Politik Universitas Riau.
Hartono, B., Eniyati, S., & Hadiono, K. (2023). Perbandingan Metode Perhitungan Jarak pada Nilai
Centroid dan Pengelompokan Data Menggunakan K-Means Clustering. Jurnal Sistem Komputer
Dan Informatika (JSON), 4(3), 503. https://doi.org/10.30865/json.v4i3.6021
Hermawan, D., Ullah, A., & Faizal, A. (2023). Rancang Bangun Keamanan Kotak Amal dengan Akses
Fingerprint Menggunakan ESP32-Cam dan Telegram Berbasis IOT. 7(3), 1013–1021.
https://doi.org/10.30865/mib.v7i3.6252
Kasus Cybercrime Dengan Studi Kasus Hacker Bjorka Terhadap Pembocoran Data Zaki Rizaldi, A.
M., Dwi Putra, R., Ul Hosnah, A., & Zaki Rizaldi, M. (n.d.). Analisis Kasus Cybercrime Dengan Studi
Kasus Hacker Bjorka Terhadap Pembocoran Data. http://jurnal.um-
tapsel.ac.id/index.php/justitia
Khairunnisa, A., Putri Efendy, M., Zamri, M., & Tusakdiyah, S. H. (n.d.). Aplikasi Pendeteksi Gejala
Penyakit Jantung Menggunakan Metode Minkowski Distance Dengan Citra Matlab (Vol. 3, Issue
1).
Mayamin, M., & Usuluddin, L. (2023). Pengaruh Absensi Sidik Jari (Finger Print) terhadap Disiplin
Kerja Pegawai. Remik, 7(1), 602–609. https://doi.org/10.33395/remik.v7i1.12091
Nishom, M. (2019). Perbandingan Akurasi Euclidean Distance, Minkowski Distance, dan Manhattan
Distance pada Algoritma K-Means Clustering berbasis Chi-Square. Jurnal Informatika: Jurnal
Pengembangan IT, 4(1), 20–24. https://doi.org/10.30591/jpit.v4i1.1253
Nuriadin, A., Dyan Nofia Harumike, Y., Tana Sanggamu, D. (2021). Sejarah perkembangan dan
implikasi internet pada media massa dan kehidupan masyarakat. SELASAR KPI: Referensi
Media Komunikasi Dan Dakwah, 1(1). https://ejournal.iainu-
kebumen.ac.id/index.php/selasar/index
Nuriadin, A. Y. D. N. (2021). Sejarah Internet di Indonesia.
Nurliza, N. N., & Widodo, E. et al. (2018). Penerapan euclidean distance pada pengenalan pola citra
sidik jarI.
Pradana, I. H. (2017). Klasifikasi citra sidik jari berdasarkan enam tipe pattern menggunakan metode
euclidean distance.
Safwandi, & Muthmainnah. (n.d.). Sistem Pendeteksi Terjemahan Kifayatul Muhtadi Ke Dalam
Bahasa Indonesia Menggunakan Metode Minkowski Distance.
Thant, A. A., & Aye, S. M. (2020). Euclidean, Manhattan and Minkowski Distance Methods For
Clustering Algorithms. International Journal of Scientific Research in Science, Engineering and
Technology, 553–559. https://doi.org/10.32628/ijsrset2073118

Identification of fingerprint image (Tulloh & Wijayanti) 77

BAMME Vol. 3 No. 2, 2023, pp. 69-78

This page is intentionally left blank.

78 10.12928/bamme.v3i2.9949