=Paper=
{{Paper
|id=Vol-3179/Short_4.pdf
|storemode=property
|title=Model of Transformation of the Alphabet of the Encoded Data as a Tool to Provide the Necessary Level of Video Image Qualityi in Aeromonitoring Systems
|pdfUrl=https://ceur-ws.org/Vol-3179/Short_4.pdf
|volume=Vol-3179
|authors=Serhii Khmelevskiy,Ivan Tupitsya,Maksym Parkhomenko,Yan Borovensky
|dblpUrl=https://dblp.org/rec/conf/iti2/KhmelevskiyTPB21
}}
==Model of Transformation of the Alphabet of the Encoded Data as a Tool to Provide the Necessary Level of Video Image Qualityi in Aeromonitoring Systems==
https://ceur-ws.org/Vol-3179/Short_4.pdf
Model of Transformation of the Alphabet of the Encoded Data
as a Tool to Provide the Necessary Level of Video Image Quality
in Aeromonitoring Systems
Serhii Khmelevskiy, Ivan Tupitsya, Maksim Parkhomenko and Yan Borovensky
Ivan Kozhedub Kharkiv National Air Force University, Sumskaya Str. 77/79, Kharkiv, 61000, Ukraine
Abstract
The drawbacks of the existing algorithms for coding data of video information resources in
the aero monitoring system are analyzed. The main ones are: the complexity of the
algorithmic implementation; loss of informative data that determine the semantic component
(carry semantic load) of the information resource. To solve these problems, a model for
transforming the alphabet of encoded data is being developed. The essence of the developed
model is to determine and take into account the significance of the elements of the encoded
data. The significance of message elements means taking into account the probabilistic
distribution of elements in the message and the structural features of the color model to
which the initial data is presented. Experimental studies are being carried out to confirm the
adequacy of the developed model. The effectiveness of the developed model is assessed from
the standpoint of ensuring the reduction of the power of the alphabet of the encoded data,
provided that the high quality of video images is ensured. The directions of further research
are determined, which involve the synthesis of the developed model with technologies for
coding data information resources without losses.
Keywords 1
transformation, information resource data, redundancy, significance, coding, aeromonitoring
1. Introduction
In the system of information support of departmental bodies, a rather important role is assigned to
the use of the aero segment [1-3]. This is due to the fact that obtaining information in real time allows
for both timely detection and prompt response to crisis situations by organizing coordinated
interaction of the relevant departmental bodies.
Accordingly, the requirements for the video information resource are increasing. The main ones
among them [4-8] are: a compact representation of the encoded data, a high level of quality of
reconstructed images in conditions of limited bandwidth of data transmission channels. For this
purpose, image coding algorithms based on conceptual approaches, which are implemented on the
basis of the JPEG platform [9-17], are quite actively used. However, it should be noted that
algorithms of the JPEG family have a number of significant drawbacks [12-27]. The main
disadvantages include the following: the complexity of the algorithmic implementation; loss of
informative data (key information) that determine the semantic component (carry semantic load) of
the information resource. Therefore, the question of finding new approaches to the compact
representation of encoded data in terms of ensuring a high level of quality becomes relevant.
The analysis of the latest scientific research indicates that the existing compression technologies
are built on the principle of multilevel processing (they assume the implementation of step-by-step
data processing) [28-34]. This means that the original alphabet is used only at the initial stage of video
Information Technology and Implementation (IT&I-2021), December 01–03, 2021, Kyiv, Ukraine
EMAIL: sserg1978@ukr.net (S. Khmelevskiy); ivan20081982@gmail.com (I. Tupitsya); maxpar76@gmail.com (M. Parkhomenko);
borovenskii2001@gmail.com (Ya.A. Borovensky)
ORCID: 0000-0001-6216-3006 (S. Khmelevskiy); 0000-0001-6806-4914 (I. Tupitsya); 0000-0001-6062-7743 (M. Parkhomenko); 0000-
0002-8747-4203 (Ya.A. Borovensky)
©️ 2022 Copyright for this paper by its authors.
Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
CEUR Workshop Proceedings (CEUR-WS.org)
311
data processing. All subsequent stages (preparation for encoding and encoding, respectively) are
associated with the processing of intermediate data. And accordingly, an increase in the number of
processing stages implies an increase in the complexity of the algorithmic implementation (which
increases the time for data processing) and a decrease in the quality of the restored images.
Therefore, it is proposed to use an approach to ensure the required level of video image quality
under conditions of limited bandwidth of data transmission channels, the essence of which is to
transform the original alphabet of the encoded data in the direction of reducing psycho-visual
redundancy by determining the significance of the encoded data elements.
Thus, the purpose of the article is to develop a model for transforming the alphabet of encoded
data to ensure the required level of video image quality in conditions of limited bandwidth of data
transmission channels in aero monitoring systems.
2. Development of a transformation model for the alphabet of encoded data
to ensure the required level of video image quality in aeromonitoring
systems
It is proposed to develop a model for transforming the alphabet of the encoded data, which will
take into account the significance of the encoded elements according to some quantitative criterion -
the sign of the element's significance. So for the message, which is given by the following expression:
X(n ) {x1; ...; x i ; ...; x n } , i 1, n , (1)
where x i - i - th message element X(n ) , i 1, n , message alphabet X(n ) will look like this:
X(m) {x1; ...; x i ; ...; x m } , i 1, m , (2)
where m - the number of elements in the alphabet X(m) messages X(n ) , i 1, m .
Taking into account the fact that at the initial stage the encoded data is represented using the RGB
model, that is, they are elements of an RGB cube, the elements of the encoded data can be represented
as follows:
x i x i (x i R , x i G , x i B ) , (3)
where x i R - the component red of the element x i ; x i G - the component green element x i ; x i B -
the component blue of the element x i .
It is proposed to transform the alphabet of the encoded data in several stages (Fig. 1). At the first
stage, the elements are determined x i messages X(n ) , which do not have a significant impact on the
semantic component of the message, that is, psycho-visual redundancy is eliminated. At this stage, the
probability distribution law is determined P ( x i ) appearance of elements x i in the message X(n ) .
1. At the next stage, the significance of the elements is determined x i messages X(n )
quantitatively. To this end, the element x i messages X(n ) represented as the center of a cube in the
RGB color space with the size 3 , i.e.:
VRGB 3 , (4)
where VRGB - the volume of the cube in the RGB color model;
- the length of the side of the cube, in pixels.
To assess the significance of elements x i messages X(n ) it is proposed to use a quantitative
feature, which is given by the following expression:
P( x i ) , (5)
S( x i )
D(x i , xi ij )
P( x )
j1
312
where S( x i ) - the coefficient of significance of the element x i messages X(n ) ;
P ( x i ) - the probability of the item appearing x i in the message X(n ) ;
x i - message element X(n ) , which is the center of the generated cube in the RGB color model;
x ij - elements of the formed color cube in the RGB model, x ij x ij (x ij , x ij , x ij ) , j 1, ,
R G B
3( 1) ;
D( x i , x ij ) - distance between element x i (the center of the formed cube) and elements x ij the
generated RGB cube.
Information resource data presented
Formation of a model for determining significance for message X(n )
using the RGB color model
elements x i
X(n ) {x1; ...; x i ; ...; x n } xi
Determination of
x i x i ( x i R , x i G , x i B ), i 1, n Determination of the law of Defining an RGB cube
probability distribution model the optimal
X(m) {x1; ...; x i ; ...; x m }, i 1, m elements in the message exponent
VRGB 3
P(x i )
Transformation of the message
and the alphabet of the encoded
data
f
X(n) X(n)
transf
S( x i )
X (n ) {x1; ...; xi ; ...; xn }
P( x i )
xi f transf (| X transf () |, S( x i )lim ) S( x i )
D(xi , xi ij )
P( x )
x i , S( x i ) S( x i )lim ;
xi j1
x ij , S( x i ) S( x i )lim , S( x ij ) S( x i )lim .
Figure 1: Structural and functional diagram of the developed model of transformation of the
alphabet of encoded data
In turn, the distance D( x i , x ij ) between element x i messages X(n ) and element x ij An RGB
color cube is defined by the following expression:
D( x i , x ij ) ( x i R x ij ) 2 ( x i G x ij ) 2 ( x i B x ij ) 2 . (6)
R G B
2. The third stage is the ranking of the data array (coefficients of significance S( x i ) elements x i
messages X(n ) ) from maximum to minimum value.
3. At the fourth stage, the significance threshold is determined S( x i ) lim elements x i messages
X(n ) (image pixels). This indicator is determined by the power of the alphabet of the reconstructed
message. Thus, taking into account the transformation of the alphabet X(m) initial data using the
proposed approach message X(n ) will take the following form:
f
X(n) X (n) , X(n ) {x1 ; ...; x i ; ...; x n } ,
t ransf
(7)
where f transf (| X transf () |, S( x i ) lim ) - function of forming a transformed message taking into
account the threshold value of the significance coefficient S( x i ) lim elements and power of the
alphabet of the transformed message;
x i - i - th element transformed message X(n ) , i 1, n ;
X transf () - transformed message alphabet X(n ) , which is formed taking into account the
significance indicator;
| X transf () | - the power of the alphabet of the transformed message X(n ) .
313
This uses the significance information (values of the coefficient of significance S( x i ) ) elements
messages X(n ) to form a transformed message X(n ) . Given this information, based on the function
f transf (| X transf () |, S( x i ) lim ) items are ranked x i transformed message X(n ) at a given level
power | X transf () | alphabet X transf () transformed message X(n ) . This is given by the following
expression:
x i f transf (| X transf () |, S( x i ) lim ) . (8)
It should be noted that it is from the power | X transf () | alphabet X transf () transformed
messages X ( n ) the degree of compression and, accordingly, the degree of distortion of the image
elements depends. Thus, at the final stage, the formation of the alphabet takes place X transf ( )
transformed messages X ( n ) , the essence of which is to replace the elements x i original message
X(n ) for which the significance coefficient S( x i ) matters less than the threshold S( x i ) lim , that is:
S( x i ) S( x i ) lim ,
elements of the alphabet X transf (m) for which the following condition is satisfied:
S( x i ) S( x i ) lim .
Thus, the process of forming a transformed message X ( n ) is given by the following expression
system:
x i , S( x i ) S( x i ) lim ;
x i (9)
x ij , S( x i ) S( x i ) lim , S( x ij ) S( x i ) lim .
In turn, the purpose of transforming the alphabet X(m) the encoded data is power reduction
| X(m) | subject to the provision of high quality video images (in terms of using compression
algorithms for coding data), i.e. the following condition must be met:
| X transf () || X(m) | . (10)
3. Evaluation of the effectiveness of the developed model of transformation
of the alphabet of encoded data from the standpoint of ensuring the
required level of video image quality in aeromonitoring systems
Analysis of transformations of the nature of the probability distribution law for the appearance of
message elements as a result of clustering based on the number of series of units.
In order to determine the optimal values of quantitative indicators (RGB cube model - length
sides of a cube, exponent ), which are used in the developed model of transformation of the alphabet
X(m) coded data, it is proposed to conduct a number of experimental studies. To determine the
effectiveness of the developed model for transforming the encoded data, a number of experiments
were carried out. Highly saturated images were used as initial data; an example of one of them is
shown in Fig. 2. The results of experimental studies to determine the optimal values of quantitative
indicators that are used in the developed model of the transformation of the alphabet X(m) of the
encoded data are shown in Figs. 3 - 4. Analysis of the results of experimental studies to determine the
optimal values of quantitative indicators that are used in the developed model of transformation of the
alphabet X(m) coded, which are presented in Fig. 3-4, indicate that:
1. The optimal value for the quantitative indicator of the RGB cube model is length side of the
cube is as follows (Fig. 3):
9. (11)
314
This is due to the fact that for a given value of the length side of the cube, the minimum
standard deviation is provided, which has the following value (Fig. 3):
CKO CKO min 0,7% .
It should be noted that this RGB cube model (with 9 ) is selected taking into account the
assessment of the complexity of the algorithmic implementation of the developed model for
transforming the encoded data (Fig. 4).
2. Analysis of the dependence of the standard deviation on the length sides of a cube for different
values of the exponent (Fig. 4) indicates that for the developed model of transformation of the
alphabet of encoded data with the optimal value of the indicator is the following:
3,5 . (12)
At 3,5 the standard deviation is provided, which has the following value (Fig. 4):
CKO 0,7 % .
Taking into account the results of the experimental studies (data from expressions (11) - (12)),
expression (5) takes the following form:
P( x i ) | X(m) |
S( x i ) . 9,55. (13)
24 | X transf () |
D(x i , x ij )3,5
P( x i )
j 1
Further, it is proposed to evaluate the effectiveness of the developed model from the standpoint of
ensuring power reduction | X(m) | alphabet X(m) of the encoded data, provided that the high quality
of video images is ensured, that is, the condition specified by expression (10) is met. So for the tested
image, which was used in the course of experimental studies, the result of applying the developed
model for transforming the alphabet of the encoded data is a decrease in the power of the alphabet by
approximately 10 times (Fig. 5):
Figure 2: Test highly saturated image
Thus, we can conclude that the proposed model for transforming the alphabet of the encoded data
allows to ensure not only the high quality of reconstructed images, but also to create more favorable
conditions for the further use of lossless compression coding technologies in order to compactly
represent the data of the information resource. Therefore, the goal of further research is to synthesize
the developed model for transforming the alphabet with technologies for compact representation of
315
encoded data without loss to increase the efficiency of delivery of data information resources while
ensuring the required level of quality.
3 MSE,%
2.5
2
1.5
1
0.5
, pix
0
0 5 10 15 20 25 30 35
7 6,6 6,2 5,8 5,4 5 4,6 4,2
3,8 3,4 3 2,6 2,2 1,8 1,4
Figure 3: Diagram the dependence of the standard deviation on the length sides of a cube for
different values of the exponent for the developed video sequence alphabet transformation models
T400 MSE,%
0.702
proc , mc
0.701
350
0.7
300
0.699
250
0.698
200 0.697
0.696
150
0.695
100
0.694
50
0.693
0 0.692
8 9 10 11 12 13 14 15 , pix
16
time
времяof processing
обработки СКО
MSE
Figure 4: Diagram of the dependence of the standard deviation and the time for data processing on
the length of the cube for the developed model of transformation of the alphabet of the video
sequence
Thus, the developed model makes it possible to ensure a sufficiently high quality of encoded data (
CKO 0,7 % ) by transforming the alphabet of the message. The essence of which is to additionally
eliminate the psycho-visual redundancy of the information resource data by determining the
significance of the elements in the message. The significance of the elements of the message is
understood as taking into account both the law of the probabilistic distribution of elements in the
message and the structural features of the color model to which the initial data are presented. This
316
makes it possible to reduce the power of the alphabet of the encoded data by almost 10 times, which,
in turn, makes it possible to create more favorable conditions for increasing the efficiency of video
data encoding from the standpoint of ensuring the required level of quality in conditions of limited
bandwidth of the data transmission channel.
35000
| X (m) |
29317
30000
25000
20000
15000
10000
5000 3070
0
RGB PM
Figure 5: Diagram of the estimation of the power of the alphabet of the encoded data for the RGB
model and the developed model
4. References
[1] Wang, S., Zhang, X., Liu, X., Zhang, J., Ma, S. and Gao, W. (2017), Utility Driven Adaptive
Preprocessing for Screen Content Video Compression. IEEE Transactions on Multimedia, Vol.
19, No. 3, pp. 660-667.
[2] Gonzales, R.C. and Woods, R.E. (2002), Digital image processing. Prentice Inc. Upper Saddle
River, 779 p.
[3] Dong, W. and Wang, J. (2016), JPEG Compression Forensics against Resizing. Computer
Science, pp. 1001-1007. DOI: 10.1109/TrustCom.2016.0168. available at:
https://www.semanticscholar.org/paper/JPEG-Compression-Forensics-against-Resizing-Dong-
Wang/620691ba5ac31d7d26f7167390ce86cb83f7a086 (accessed (accessed 05 June 2021).
[4] Richter, T. (2017), Error Bounds for HDR Image Coding with JPEG XT. Data Compression
Conference (DCC), pp. 122-130. DOI: 10.1109/DCC.2017.7.
[5] Xiao, W., Wan, N.A., Hong and Chen, X. (2020), A Fast JPEG Image Compression Algorithm
Based on DCT. IEEE International Conference on Smart Cloud (SmartCloud), pp. 106-110. DOI:
10.1109/ SmartCloud49737. 2020.00028.
[6] Rippel, O. (2019), Learned Video Compression. IEEE/CVF International Conference on
Computer Vision (ICCV), pp. 3453-3462. DOI: 10.1109/ICCV. 2019.00355.
[7] Bienik, J., Uhrina, M., Kuba, M. and Vaculik, M. (2016), Performance of H.264, H.265, VP8 and
VP9 Compression Standards for High Resolutions. 19th International Conference on Network-
Based Information Systems (NBiS), pp. 246-252. DOI: 10.1109/NBiS. 2016.70.
[8] Wang, X., Xiao, J., Hu, R. and Wang, Z. (2017), Cruise UAV Video Compression Based on
Long-Term Wide-Range Background. Data Compression Conference (DCC), pp. 466-467. DOI:
10.1109/DCC.2017.71.
[9] Minallah, N., Gul, S. and Bokhari, M. (2015), Performance Analysis of H.265/HEVC (High-
Efficiency Video Coding) with Reference to Other Codecs. 13th International Conference on
Frontiers of Information Technology (FIT), pp. 216-221. DOI: 10.1109/ FIT.2015.46.
317
[10] Djelouah, A., Campos, J., Schaub-Meyer, S. and Schroers, C. (2019), Neural Inter-Frame
Compression for Video Coding. IEEE/CVF International Conference on Computer Vision
(ICCV), pp. 6420-6428. DOI: 10.1109/ICCV.2019. 00652.
[11] Narmatha, C., Manimegalai, P. and Manimurugan, S. (2017), A LS-compression scheme for
grayscale images using pixel based technique. International Conference on Innovations in Green
Energy and Healthcare Technologies (IGEHT), pp. 1-5, DOI: 10.1109/ IGEHT.2017.8093980.
[12] Poolakkachalil, T.K., Chandran, S., Muralidharan, R. and Vijayalakshmi, K. (2016),
Comparative analysis of lossless compression techniques in efficient DCT-based image
compression system based on Laplacian Transparent Composite Model and An Innovative
Lossless Compression Method for Discrete-Color Images. 3rd MEC International Conference on
Big Data and Smart City (ICBDSC), pp. 1-6, DOI: 10.1109/ICBDSC.2016.7460360.
[13] Wang, Z., Liao, R. and Ye, Y. (2020), Joint Learned and Traditional Video Compression for P
Frame. IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops
(CVPRW), pp. 560-564. DOI: 10.1109/CVPRW50498.2020.00075.
[14] Bui, V., Chang, L., Li, D., Hsu, L. and Chen, M. (2016), Comparison of lossless video and image
compression codecs for medical computed tomography datasets. IEEE International Conference
on Big Data (Big Data), pp. 3960-3962. DOI: 10.1109/BigData. 2016.7841075.
[15] Akbari, M., Liang, J., Han, J. and Tu, C. (2020), Learned Variable-Rate Image Compression
With Residual Divisive Normalization. IEEE International Conference on Multimedia and Expo
(ICME), pp. 1-6. DOI: 10.1109/ICME46284.2020.9102877.
[16] Shinde, T. (2019), Efficient Image Set Compression. IEEE International Conference on Image
Processing (ICIP), pp. 3016-3017, DOI: 10.1109/ICIP. 2019.8803230.
[17] Lin, J., Liu, D., Li, H. and Wu, F. (2020), M-LVC: Multiple Frames Prediction for Learned
Video Compression. IEEE/CVF Conference on Computer Vision and Pattern Recognition
(CVPR), pp. 3543-3551. DOI: 10.1109/CVPR42600. 2020.00360.
[18] Artuğer, F. and Özkaynak, F. (2018), Fractal Image Compression Method for Lossy Data
Compression. International Conference on Artificial Intelligence and Data Processing (IDAP),
pp. 1-6. DOI: 10.1109/IDAP. 2018.8620735.
[19] Arnob, P., Tanvir, Z., Prajoy, P., Rafi, A., Muktadir Rahman, M. and Mamdudul Haque, Kh.
(2015), Iris image compression using wavelets transform coding. 2nd International Conference
on Signal Processing and Integrated Networks (SPIN), pp. 544-548,
DOI: 10.1109/SPIN.2015.7095407.
[20] Zhu, X., Liu, L., Jin and Na Ai, P. (2016), Morphological component decomposition combined
with compressed sensing for image compression. IEEE International Conference on Information
and Automation (ICIA), DOI: 10.1109/ICInfA.2016.7832096.
[21] Wang, S., Kim, S. M., Yin, Z., and He, T. (2017), Encode when necessary: Correlated network
coding under unreliable wireless links. ACM Transactions on Sensor Networks, Vol. 13(1). DOI:
10.1145/ 3023953.
[22] Phatak, A. (2016), A Non-format Compliant Scalable RSA-based JPEG Encryption Algorithm.
International Journal of Image. Graphics and Signal Processing, Vol. 8, No. 6, pp. 64-71. DOI:
10.5815/ijigsp. 2016.06.08.
[23] Barannik, V., Tupytsya, I., Parkhomenko, M., Azatov, A., Pershin, A., Gurzhii, P., Shaikhanova,
A. and Karpiinski, M. (2020), The concept of a quantitative sign formation for the internal
restructuring of information resource data. Przetwarzanie, transmisja i bezpieczeństwo
informacji', pp. 41-52.
[24] Wu, H., Sun, X., Yang, J., Zeng, W. and Wu, F. (2016), Lossless Compression of JPEG Coded
Photo Collections. IEEE Transactions on Image Processing, Vol. 25, No. 6, pp. 2684-2696. DOI:
10.1109/ TIP.2016.2551366.
[25] Chen, C. and Zhuo, Y. (2017), A research on anti-jamming method based on compressive
sensing for OFDM analogous system. IEEE 17th International Conference on Communication
Technology (ICCT), pp. 655-659, DOI: 10.1109/ICCT.2017. 8359718.
[26] Wang, S., Kim, S., Yin, Z. and He, T. (2015), Encode when necessary: Correlated network
coding under unreliable wireless links. ACM Transactions on Sensor Networks, Vol. 13, No.1,
pp. 24-29, DOI: 10.1145/ 3023953.
318
[27] Han, S., Mao, H. and Dally, W. (2015), Deep compression: Compressing deep neural networks
with pruning, trained quantization and huffman coding, available at:
https://webcache.googleusercontent.com/search?q=cache:DSoAo2aheUoJ:https://arxiv.org/pdf/1
510.00149+&cd=2&hl=uk&ct=clnk&gl=ua (accessed 05 June 2021).
[28] Zhurakovskyi, B., Boiko, J., Druzhynin, V., Zeniv, I. and Eromenko, O. (2020), Increasing the
efficiency of information transmission in communication channels. Indonesian Journal of
Electrical Engineering and Computer Science, Vol. 19(3), pp. 1306-1315. DOI:
10.11591/ijeecs.v19.i3.
[29] Barannik, V., Tupitsya, I., Dodukh, O., Barannik, V. and Parkhomenko, M. (2019), The Method
of Clustering Information Resource Data on the Sign of the Number of Series of Units as a Tool
to improve the Statistical Coding Efficiency. IEEE 15th International Conference on the
Experience of Designing and Application of CAD Systems (CADSM), pp. 32-35.
DOI: 10.1109/CADSM.2019.8779243.
[30] Barannik, V., Tupitsya, I., Gurzhii, I., Barannik, V., Sidchenko, S. and Kulitsa, O. (2019), Two-
Hierarchical Scheme of Statistical Coding of Information Resource Data with Quantitative
Clustering. IEEE International Conference on Advanced Trends in Information Theory (ATIT),
pp. 89-92. DOI: 10.1109/ATIT49449.2019.9030451.
[31] Barannik, V., Tupitsya, I., Barannik, V., Shulgin, S., Musienko, A., Kochan, R. and Veselska, O.
(2019), The Application of the Internal Restructuring Method of the Information Resource Data
According to the Sign of the Number of Series of Units to Improve the Statistical Coding
Efficiency. 10th IEEE International Conference on Intelligent Data Acquisition and Advanced
Computing Systems: Technology and Applications (IDAACS), pp. 65-69.
DOI: 10.1109/IDAACS.2019.8924460.
[32] Barannik, V., Tupitsya, I., Kovalenko, O., Sidchenko, Y., Yroshenko, V. and Stepanko, O.
(2020), The analysis of the internal restructuring method efficiency used for a more compact
representation of the encoded data. Advanced Trends in Information Theory (ATIT’2020):
proceedings of the Intern. Conf., pp. 48-51. DOI: 10.1109/ATIT50783.2020.9349307.
[33] Yudin, O., Artemov, V., Krasnorutsky, A., Barannik, V. Tupitsya, I. and Pris, G. (2021),
Creating a mathematical model for estimating the impact of errors in the process of
reconstruction of non-uniform code structures on the quality of recoverable video images.
Advanced Trends in Information Theory (ATIT’2021): proceedings of the Intern. Conf., pp. 38-
41. DOI: 10.1109/ATIT54053.2021.9678887.
[34] Khmelevskiy, S., Tupitsya, I., Mahdi, Q. A., Musienko О., Parkhomenko, M., & Borovensky , Y.
(2021). Development of the external restructuring method to increase the efficiency of
information resource data encoding. Information Processing Systems, 3(166), pp. 52-61. DOI:
10.30748/soi.2021.166.06.
319