Encryption of Dynamic Areas of Images in Video based on Certain

Geometric and Color Shapes

NASHAT AL BDOUR

Computer and Communication Engineering,

Tafila Technical University,

Tafila,

JORDAN

Abstract: - The paper is devoted to the search for new approaches to encrypting selected objects in an image.

Videos were analyzed, which were divided into frames, and in each video frame, the necessary objects were

detected for further encryption. Images of objects with a designated geometric shape and color characteristics of

pixels were considered. To select objects, a method was used based on the calculation of average values, the

analysis of which made it possible to determine the convergence with the established image. Dividing the selected

field into subregions with different shapes solves the problem of finding objects of the same type with different

scales. In addition, the paper considers the detection of moving objects. The detection of moving objects is carried

out based on determining the frame difference in pixel codes in the form of a rectangular shape. Cellular automata

technology was used for encryption. The best results were shown by the transition rules of elementary cellular

automata, such as: 90, 105, 150, and XOR function. The use of cellular automata technologies made it possible to

use one key sequence to encrypt objects on all video frames of the video. Encryption results are different for the

same objects located in different places of the same video frame and different video frames of the video sequence.

The video frame image is divided into bit layers, the number of which is determined by the length of the code of

each pixel. Each bit layer is encrypted with the same evolution, which is formed by one initial key bit sequence.

For each video frame, a different part of the evolution is used, as well as for each detected object in the image.

This approach gives different results for any objects that have a different location both on the video frame image

and in different video frames. The described methods allow you to automate the process of detecting objects on

video and encrypting them.

Key-Words: - encryption, image, cellular automata, key array, image section, object image detection, evolution,

Wolfram's rule.

Received: May 15, 2022. Revised: January 23, 2023. Accepted: February 19, 2023. Published: March 29, 2023.

1 Introduction

In modern society, information technology has

reached such a stage of development that video is

used in almost all areas of human activity. Video

surveillance systems of various structures and various

functional purposes are widely used. Now

information about various events in most cases is

presented using video. However, there are cases when

the video contains various prohibited content that

occupies insignificant areas in the field of the video

display. Such content can be images: cigarettes,

alcohol, naked body, prohibited signs, and others. The

presence of prohibited images imposes a ban on all

videos. In this case, the video can be shown, and

prohibited areas can be hidden (filled with one color,

for example, white). Such areas are usually defined by

the user and hidden by the user. Existing approaches

are based on the formation of another video with

hidden image areas. This is necessary to save the

original image, which results in the use of additional

memory. There are also areas of the image on the

video that carry confidential or secret information.

Such areas of the image must be encrypted. At the

same time, the encryption-based approach does not

require additional memory, since it can be determined

using the encryption and decryption key and does not

WSEAS TRANSACTIONS on INFORMATION SCIENCE and APPLICATIONS

DOI: 10.37394/23209.2023.20.13

Nashat Al Bdour

E-ISSN: 2224-3402

109

Volume 20, 2023

require the formation of additional video.

One of the main functions for hiding parts of the

video image is the function of searching and selecting

the specified images that may appear in some video

frames during the entire video. Searching for such

areas in the image by the user leads to a lot of time.

To automate the process and increase the

reliability of the selection of specified areas, it is

necessary to use methods for automatically detecting

such areas in each video frame for further encryption.

The objective of the research is to increase the

reliability of the selected areas and reduce the time

spent on processing all video frames based on

methods that use process automation. Also, the task is

to detect dynamically changing objects in the video

(moving, resizing, etc.), which are fixed using

rectangular selection windows. Selected forms on

each video frame are encrypted using cellular

automata (CA) technologies, which allows solving the

problem of using the optimal size encryption key

when encrypting large amounts of video data.

All modern works aimed at creating methods for

detecting and selecting a graphic object by geometric

and color structure are divided into two large groups,

[1], [2]: generalizing and distinguishing. Generalizing

methods are based on building a given model of the

image of the detected object and assessing the

accuracy of the model of the new image found in the

visual picture based on this model. The most popular

generalizing methods are, [2], [3]: random field

model, [4], [5], implicit form model, [6], constellation

model, [7]. The main characteristics of the methods

using these models are presented in [2].

The second group of methods based on the

difference is that a known classifier is created or used,

with the help of which the differences between the

negative and positive images of a particular training

sample are determined. The best-known difference-

based methods are the Viola-Johnson, [8], LeCun, [9],

and Papageorgiou, [10], methods. Comparative

characteristics of these methods are presented in the

table in [2].

Both groups of methods and each of the methods

in each specific case of solving a specific problem are

the most effective. Influencing factors are:

performance, memory size, implementation

complexity, and other parameters. Systems that

implement methods for selecting objects in the visual

picture use the training mode and the mode of direct

search and selection of an object.

In [11], the selection of objects is based on the

perception and action in a three-dimensional scene

from different points of view. The results obtained in

this work allow you to select objects in a complex

visual environment.

Among the works devoted to image encryption in

this paper, attention is paid to partial image

encryption on each video frame of the video. In

general, all existing methods, [12], [13], [14], [15],

[16], [17], [18], that are aimed at image encryption

can be used to encrypt parts of images. The analysis

of works devoted to image encryption was carried out

in [19], [20], [21]. The works, [22], [23], describe

methods for encrypting parts of an image. These

methods use selective encryption of the specified

image area based on unique attributes, such as image

frequency, total brightness, special compression area,

etc., [23], [24], [25], [26], [27], [28].

The most promising methods for partial image

encryption are methods using cellular automata

technologies, [19], [21], [29], [30]. This technology

makes it possible to reduce the key length, as well as

to reduce the time it takes to encrypt image sections in

all frames of the analyzed video.

2 Method for Searching for

Homogeneous Objects on Video Frames

The simplest method is to enumerate the numerical

values of the codes of each pixel in the image of each

video frame for subsequent comparison of the

sequence of numbers with the reference sequence that

must be selected on the image of each frame.

However, this method is time-consuming. In this

method, small differences in codes can lead to false

results that make it impossible to select the desired

area.

To select an area on an image, this paper uses the

method of calculating the average value for the entire

selected area and for its individual sections, [31]. In

this case, the average values of all selected areas must

be strictly ordered. The average value is calculated for

all values of the pixel codes belonging to the image

area of a given size, and the allowable interval is set,

within which the image area is determined by the

given average value. Such a method is described in

WSEAS TRANSACTIONS on INFORMATION SCIENCE and APPLICATIONS

DOI: 10.37394/23209.2023.20.13

Nashat Al Bdour

E-ISSN: 2224-3402

110

Volume 20, 2023

detail in the paper, [31], for selecting the human

auricle in the image.

It is necessary to take into account the size of the

image area to be selected. The size of the image area

can be set in advance by the user. In this case, a false

selection of an image area of a given size is possible.

An example of such a false selection is shown in

Figure 1.

Fig. 1: An example of a false selection of an image

area of a given size based on the average value.

Figure 1 shows that the images are different, but

they have the same average values (the difference is

0.97018888). In this case, the image structure of the

selected area is not determined. Only one quantitative

value is determined, which does not carry information

about the structure of the image and the distribution

of color and brightness characteristics in it.

The structure of the image and the distribution of

color and brightness characteristics are determined by

dividing the selected area into smaller areas, and the

necessary selection of such areas is also carried out.

Options for splitting the images are shown in Figure 1

and on a smaller area in Figure 2.

Fig. 2: Options for partitioning the images shown in

Fig. 1, on a smaller area.

In Figure 2, the partitioning options are shown

with a different number of subareas and with different

locations. As can be seen from Figure 2, different

geometric shapes fall into each field, which

accordingly leads to different average values. For the

left image (Figure 2), the average values can be

represented as a matrix

and for the right image (Figure 2) the matrix of

averages has the following values

Accordingly, the difference between such average

values for each selected area shows that the images

are different. The matrix of difference values has the

following form

In this case, four rectangular regions are used, so

the arrays of averages can be represented as two-

dimensional arrays, which simplifies the processing

of the obtained data.

In the case of using video in each video frame,

image fragments may be distorted. This is due to the

video filming, the distance to the video camera, etc.

Changing the distance to the video camera entails a

change in the scale of both the entire image and

individual sections of the images of each video frame.

In such situations, controlled areas of video with

the same picture elements and different scales cannot

be determined using average values. Therefore, to

identify images in each video frame, a matrix of ratios

of the average values of all sub-areas into which the

controlled area of images is divided is used

The relationship matrix R has a size of n×m

(where n is the number of sub-areas along the X-

coordinate and m is the number of sub-areas along the

Y-coordinate). The more sub-areas, the higher the

accuracy of image identification. However, the

processing speed is reduced, which is unacceptable

for processing video sequences in real-time. In

addition, to process the quantitative values of the

relationship matrix R, it is necessary to use special

algorithms and form a base of reference matrices Rref.

However, the construction of such a matrix is possible

for rectangular areas in the image.

If it is necessary to select images of certain

geometric shapes with specified color and brightness

characteristics, then the formation of different shapes

WSEAS TRANSACTIONS on INFORMATION SCIENCE and APPLICATIONS

DOI: 10.37394/23209.2023.20.13

Nashat Al Bdour

E-ISSN: 2224-3402

111

Volume 20, 2023

of subareas and their number is carried out. Figure 3

shows different options for the shape of sub-regions

for different images.

Fig. 3: Variants of the forms of sub-areas into which

controlled areas of images are divided for different

images

Figure 3 shows that different shapes and different

numbers of sub-areas can be used for different images

(separated by light lines in each image). For each

option, a sequence vector of quantitative values

obtained from the analysis of each subarea is formed.

In general, the vector structure has the form

where - average value for i - th subarea.

In this case, for each variant, a rigid arrangement

of the sequence of numbers in the vector is set. To

sequentially compare the numbers inside the vector

with the reference ones, the first ones should be the

average values of those sub-areas that carry the

greatest amount of information about the entire

selected area. If there are scale changes on the image,

then a matrix of numbers is formed that corresponds

to the relationship between all average values of the

vector of average values. In this case, subareas are

also formed taking into account large-scale changes in

the entire selected area of the video frame image. In

accordance with scale changes and different shapes of

images of the selected areas of the image, the

relationship vectors also have different sizes.

Based on the considered approaches, an important

task is to find the optimal shapes and location of

subdomains that provide the most information in the

analysis of the selected area.

3 Detection of Moving Objects in the

Visual Scene

In many areas, some tasks are aimed at detecting and

hiding moving objects on the video scene. To solve

such problems, the video is divided into a sequence of

video frames, each of which is an image, and if there

is movement in the video, then the images of the

video frames are different. At the same time, pictures

can change completely in adjacent video frames on

the video. In this case, a large number of pixels are

allocated that have changed their codes in two

adjacent video frames. In practice, the entire image in

each subsequent video frame can be selected. Among

such changes in the entire background of the image, it

is very difficult to distinguish a moving object.

This work is aimed at detecting not only moving

objects on a weakly changing video background but

also detecting moving objects that have certain

geometric shapes. If the background of the entire

image does not change, but only the analyzed object

changes (its location changes), then the solution to

this problem is simplified. If the general background

in adjacent video frames of the video changes, then

the solution to the object selection problem becomes

more complicated, since there is an additional task of

preliminary identification of the moving area of the

video frame image. And in this case, the preliminary

capture of the moving image area is not always used.

To select a moving object without significant

changes in the background, this work uses an

algorithm for calculating the difference in the codes

of two pixels of adjacent video frames located on the

same coordinates. This method is described in detail

in [32], [33]. The method is quite simple and is

widely used in solving various problems of image

processing. A pixel-by-pixel comparison of codes in

two adjacent video frames is carried out. Pixels are

determined that, compared to pixels in previous

frames, have changed their color and brightness code.

Selected pixels capture motion on the analyzed video.

All other pixels are defined as pixels belonging to the

background of the image.

The selection of the moving area is carried out by

selecting a rectangular area, which is determined by

the selected pixels (Figure 4).

WSEAS TRANSACTIONS on INFORMATION SCIENCE and APPLICATIONS

DOI: 10.37394/23209.2023.20.13

Nashat Al Bdour

E-ISSN: 2224-3402

112

Volume 20, 2023

Fig. 4: An example of selecting an area of a selected

object using the algorithm for determining the

difference in pixel codes in adjacent video frames

Figure 4 shows a rectangular area describing the

moving object. The yellow robot movement sequence

starts from the top (initial video frame). The right side

of the highlighted area indicates where the robot was

in the previous video frame. This example shows that

the size of the selected area can change. It all depends

on other transformations of the selected area (rotation,

scaling, etc.). A different number of pixels are

selected for each frame.

Also, to detect moving objects, the algorithm for

superimposing several adjacent frames of the video

stream can be used. In this case, the difference in

pixel codes is calculated or a bitwise XOR operation

is performed for all pixels. All matched bits and codes

become black, and non-matched bits form codes of

other colors that form the selected pixels.

Another thing is when the background of the

whole picture in neighboring video frames changes

completely and more than 90% of the pixels are

highlighted (according to the algorithm discussed

above) as participating in motion. In this case, the

most appropriate selection algorithm is an algorithm

that uses the calculation of average values and the

construction of a matrix of relations that allows you to

identify the desired area of the image. Using an

algorithm based on the calculation of average values

allows us to select stationary areas.

In addition, this approach allows the selection of

areas with the same geometric shape, but different

colors (codes) of pixels (Figure 5).

Fig. 5: Examples of images of the same geometric

shapes, but with different colors

To solve the problem, the image is preliminarily

binarized (Figure 6) and the necessary average values

are calculated with division into subareas. In binary

images, it is easier to determine the average values

and set the shapes of the subareas.

Fig. 6: Examples of binarization of images presented

in Fig. 5. The brightness level during binarization is

set to 50%

WSEAS TRANSACTIONS on INFORMATION SCIENCE and APPLICATIONS

DOI: 10.37394/23209.2023.20.13

Nashat Al Bdour

E-ISSN: 2224-3402

113

Volume 20, 2023

After converting the original image into a bitmap,

the subarea shapes are selected, the average values are

calculated, and the relationship matrix is formed.

4 Encryption of Selected Areas of the

Image

To encrypt selected areas on video, there is a need to

encrypt such a section in each video frame. For this,

one encryption key can be used for the image of each

video frame. However, if the image of the selected

area in each video frame has the same geometric and

luminance-color structure, then the encryption result

will also be represented by the same image. This

approach to encrypting video sequences is not always

acceptable. For example, several different areas can

be selected in one visual picture. Accordingly, the

video will display different selected areas with the

same encryption results.

To solve the problem of separating encryption

areas by encryption keys, as well as by encryption

methods, it is necessary to use additional methods for

their specific and effective application. Since stream

encryption is most often used, it is enough to separate

the key pseudo-random bit sequence and use each part

of it as a key sequence for each selected image area in

each video frame.

One of the options for such encryption is the use

of a pseudo-random number generator based on CA

with active cells, [29], [30], [31], [32], [33], [34].

Such a method is described in [20]. This paper shows

that it is enough to use the three most significant bits

in each color byte of the image for encryption.

However, although this approach gives high image

quality, it is time-consuming and not always

acceptable.

To reduce the time spent when encrypting

individual sections of the sequence, it is advisable to

use the methods described in [19], [20], [21]. In these

works, CAs are also used to generate the encryption

key. In this case, the original key sequence has a

length equal to the dimension of the video frame

(vertical or horizontal image). The initial key

sequence determines the initial state of the elementary

CA (ECA).

Based on the initial states of the ECA and the

selected transition rule, a two-dimensional evolution

is formed, which coincides in dimension with the

image size of the frame of the analyzed video

sequence. An example of encryption of selected areas

on the image of one video frame in Figure 7 is shown.

Fig. 7: An example of encryption of selected areas on

the image of one video frame using two-dimensional

ECA evolution. The transition function of 105 ECA is

used with a shift by one step of evolution

Figure 7 shows the original image of a video

frame with selected areas (on the left) and the same

image with encrypted areas that are selected in the

original image. The image shows that the selected

areas have different color and brightness structures. In

this case, the original image has the same selected

areas.

The process of encrypting the selected areas of

the image is to perform the following functional steps.

1. The video into sequences of video frames is

divided.

2. In the image of each video frame, a given area

of the image is selected.

3. The encryption of the selected areas using the

selected encryption method is carried out.

4. A new video is formed from a sequence of

video frames with encrypted image sections.

The first stage is implemented using a variety of

software and hardware video recording and video

generation.

The resulting sequence of video frames is

analyzed in the second stage in order to select the

necessary image areas present in the images of each

video frame. Such areas can be rectangular areas that

WSEAS TRANSACTIONS on INFORMATION SCIENCE and APPLICATIONS

DOI: 10.37394/23209.2023.20.13

Nashat Al Bdour

E-ISSN: 2224-3402

114

Volume 20, 2023

highlight moving objects, as well as areas with a

certain geometric and color structure. To select each

of these areas, methods that allow you to do this were

previously considered.

To encrypt selected areas (third stage), the

necessary time costs are taken into account. If the

number of video frames is large, then the use of the

method of generating a pseudo-random key gamut

leads to a large time expenditure. Although the

method showed high reliability of encryption. The

most acceptable is the method that uses CA to form a

key array, [19], [20], [21]. The method shows

particular efficiency in cases where there are several

different selected areas on the images of video

frames. The choice of the encryption method also

takes into account the fact that the encrypted areas of

the images should differ in color structure for the

images of all video frames.

In the fourth stage, a video is formed in which

there are encrypted areas.

To encrypt the selected areas of the image, a

method is used that forms a key array based on the

CA. Rules are used: 90, 105, 150, and XOR functions,

[21], examples of which in Figure 8 are shown.

Fig. 8: Examples of key arrays formed based on ECA

evolution using rules 90, 105, 150, and XOR

functions. The XOR function uses the analysis of six

neighboring cells (three at each step of evolution) in

the previous two steps of evolution

Sections of key arrays, which coincide in a

location with the selected sections on the images of

video frames, are the corresponding selected key

arrays for each selected area. Since the key arrays

differ throughout the field of the created evolution,

the encryption result for each selected area is also

different. This situation allows us to hide the fact that

the same information is hidden in each video frame.

Examples of encryption of identical selected areas

with different locations on the image of one video

frame in Figure 9 are shown, and examples of

encryption of different areas using the same key array

in Figure 10 are shown.

Fig. 9: An example of encryption of identical selected

areas with different locations on the image of one

video frame. Rule 105 was used with a shift by one

step of evolution

Fig. 10: An example of encryption of different

selected areas on the image of one video frame. The

XOR function was used, taking into account

neighboring cells of the two previous steps of

evolution

As can be seen from Figure 9 and Figure 10, the

structures of images obtained as a result of encryption

are different, which determines the high quality of

encryption.

If you use the same key array to encrypt all video

frames, then for stationary selected areas, the

encryption results in all video frames being the same.

If dynamically changing areas or areas that change

their location are selected, then the encryption results

of such areas differ.

Improving the quality of encryption is achieved

by using different key arrays for each video frame.

The formation of such arrays can be achieved by

various methods. One such method is to use different

ECA transition rules that give the best encryption

quality. Such rules are studied and described in [21].

WSEAS TRANSACTIONS on INFORMATION SCIENCE and APPLICATIONS

DOI: 10.37394/23209.2023.20.13

Nashat Al Bdour

E-ISSN: 2224-3402

115

Volume 20, 2023

The use of different rules allows you to form different

key arrays on the same initial states of the ECA.

However, a large number of video frames does not

allow using a separate ECA transition rule with the

same initial settings for each video frame. Therefore,

the paper proposes to use an approach based on

increasing the number of evolutionary steps, which

allows increasing the key array and using its

fragments to encrypt the image of each video frame.

This approach allows you to use only one key

sequence for the entire video. This key sequence is

bit-based and forms the initial states of the ECA used

to form the corresponding evolution. The key

sequence can be represented in either decimal or

binary codes. An example of such video encryption

on individual video frames in Figure 11 is shown.

Fig. 11: An example of encryption of selected areas

based on one key sequence and different key arrays

for each video frame

In Figure 11 are stationary selected areas with

different encryption results in each video frame.

An approach was used to form key arrays, which

consists of the formation of one initial key sequence.

All bit layers into which the image of the video frame

was divided were encrypted using the generated key

arrays. The key array was formed using the formation

of ECA evolution according to a given transition rule.

For each bit layer, the key array began with different

evolution time steps. In the same way, the formation

of key arrays for the bit layers of the image of each

video frame was carried out. In fact, only one initial

key sequence is chosen, and different "shifted" key

arrays are used for encryption, which are taken from

the generated evolution. Combinations of different

transition rules can also be used.

One key sequence represents ECA early in

evolution. As a rule, its length corresponds to the

dimension of the image. From one key sequence,

many evolutions can be formed, which is determined

by the number of ECA transition rules used according

to Wolfram. Each image of a video frame is divided

into binary layers, the number of which is equal to the

length of the code (number of bits) that goes into

encoding the color characteristics of each pixel. If

three bytes are used to encode one pixel (1 byte is the

red code, 1 byte is the green code, and 1 byte is the

blue code), then the image is divided into 24 binary

layers.

The generated evolution (key two-dimensional

array) has a higher dimension than the binary layer.

An evolution fragment is used to encrypt each binary

layer. As a rule, each fragment of evolution for each

subsequent binary layer begins with each subsequent

time step of evolution. This will make it possible to

use different key arrays to encrypt objects in different

video frames.

This approach allows you to get different images

after encryption for the same image sections both in

one and in different video frames.

5 Conclusion

The paper considers the process of detecting objects

on frames of a video sequence to encrypt them. Using

a method based on calculating the average values of

the selected area and dividing the area into sub-areas

allows selecting of objects of a given geometric

structure and color range, as well as selecting the

same shapes, changed to scale. An experiment was

conducted that showed that cellular automata are best

suited for selecting and encrypting images of identical

objects with different locations. Cellular automata

make it possible to form a high-dimensional evolution

that gives different key arrays for all selected objects

in video frames. The initial key array is a bit sequence

indicating the initial state of the elementary cellular

automata. The rules that give high-quality encryption

(90, 105, 150, and XOR function based on two

previous evolution steps) are defined. In this case,

only one rule can be used to encrypt all video frames.

WSEAS TRANSACTIONS on INFORMATION SCIENCE and APPLICATIONS

DOI: 10.37394/23209.2023.20.13

Nashat Al Bdour

E-ISSN: 2224-3402

116

Volume 20, 2023

In further research, the author plans to use

cellular automata that implement the detection of

different and identical objects on video. The research

will also focus on the use of two-dimensional cellular

automata.

References:

[1] Roth, P.M. Survey of Appearance-Based

Methods for Object Recognition / P.M. Roth,

M. Winter // Technical Report ICG-TR-01/08,

Institute for Computer Graphics and Vision,

Graz University of Technology, Austria,

January, 2008. – p. 68

[2] A.N. Alfimtsev, I.I. Lychkov. The Technique of

Real-Time Object Detection in the Video

Stream. Вестник ТГТУ. 2011. Том 17. № 1.

Transactions TSTU. pp. 44-55.

[3] Murphy, K.P. Models for Generic Visual

Object Detection / K.P. Murphy // Technical

report, Department of Computer Science,

University of British Columbia, Vancouver,

Canada, May, 2005. – p. 8

[4] Stan Z. Li, “Markov Random Field Modeling in

Image Analysis”, Computer Science

Workbench, Springer, 2001. p. 323

[5] Gimel'farb G.L., “Image Textures and Gibbs

Random Fields”, Kluwer Academic Publishers:

Dordrecht, 1999. p. 250

[6] Greenspan, M. Efficient Tracking with the

Bounded Hough Transform / M. Greenspan, L.

Shang, P. Jasiobedzki // Proceedings of

Computer Vision and Pattern Recognition. –

2004. – Vol. 1. – p. 520–527.

[7] Fergus, R. Object Class Recognition by

Unsupervised Scale-Invariant Learning / R.

Fergus, P. Perona, A. Zisserman // Proceedings

of Computer Vision and Pattern Recognition. –

2003. – Vol. 2, Sec. II. – p. 264–271.

[8] Viola, P. Robust Real-Time Object Detection /

P. Viola, M. Jones // Intl. J. Computer Vision. –

2004. – Vol. 57(2). – p. 137–154.

[9] LeCun, Y. Learning Methods for Generic

Object Recognition with Invariance to Pose and

Lighting / Y. LeCun, F.-J. Huang, L. Bottou //

Proceedings of Computer Vision and Pattern

Recognition. – 2004. – Vol. 2. – p. 97–104.

[10] Papageorgiou, C.A Trainable System for Object

Detection / C. Papageorgiou, T. Poggio // Intl. J.

Computer Vision. – 2000. – Vol. 38(1). – p. 15–

33.

[11] Mela Francisco. Active Object Detection: An

Active Perception Approach to Object

Detection. Thesis for: MSC Mechatronics

Engineering. Advisor: Serhan Cosar

[12] Stepan Bilan, Andrii Demash. High

performance encryption tools of visual

information based on cellular automata. -

Information Technology and Security. - 2016. -

Vol. 4, № 1(6). - С. 62-75.

[13] Xingyuan Wang, Dapeng Luan. (2013). A novel

image encryption algorithm using chaos and

reversible cellular automata. Communications

in Nonlinear Science and Numerical Simulation

18(11):3075–3085

[14] A.L.A. Dalhoum, A. Madain, H. Hiary, Digital

image scrambling based on elementary cellular

automata, Multimedia Tools Appl. 75 (24)

(2016) 17019–17034.

[15] T.H. Chen, M. Zhang, J.H. Wu, C. Yuen, Y.

Tong, Image encryption and compression based

on kronecker compressed sensing and

elementary cellular automata scrambling, Opt.

Laser Technol. 84 (2016) 118–133.

[16] Wassim Alexan, Mohamed khaled Elbeltagy.

(2021). Lightweight Image Encryption: Cellular

Automata and the Lorenz System. Conference:

2021 International Conference on

Microelectronics (ICM): 34-39.

[17] Juan Contreras , Marco Ramírez , Jesús

Aboytes. Image Encryption System Based on

Cellular Automata and S-Box. Research in

Computing Science 148(10), 2019. pp. 153–161

[18] Wang Y, Zhao Y, Zhou Q, et al. (2018) Image

encryption using partitioned cellular automata.

Neurocomputing 275:1318–1332

[19] Nashat Albdour, Image encryption

methodology based on cellular automata,

Journal of Theoretical and Applied Information

Technology, 2022. Vol.100. No 23.

[20] Nashat Albdour and Ayman M Mansour

Optimal steganographic protection method

based on image encryption, Przegląd

Elektrotechniczny, 2021, Vol.97. No 6.

[21] Nashat Albdour,Selection of Intelligent Rules

for the Evolution of Elementary Cellular

Automata for Image Encryption, WSEAS

Transactions on Systems and Control 2022,

Vol.17.

WSEAS TRANSACTIONS on INFORMATION SCIENCE and APPLICATIONS

DOI: 10.37394/23209.2023.20.13

Nashat Al Bdour

E-ISSN: 2224-3402

117

Volume 20, 2023

[22] Marc Van Droogenbroeck. Partial encryption of

images for real-time applications. Fourth IEEE

Signal Processing Symposium

[23] Wonyoung Jang. Sun-Young Lee. Partial image

encryption using format-preserving encryption

in image processing systems for Internet of

things environment. International Journal of

Distributed Sensor Networks 2020, Vol. 16(3).

pp. 1-17.

[24] Kumari S. A research paper on cryptography

encryption and compression techniques. Int J

Eng Comput Sci 2017, 6(4): 20915–20919.

[25] Kaur M and Kumar V. A comprehensive review

on image encryption techniques. Arch Comput

Method Eng 2018, 27: 15–43.

[26] Younes MAB. A survey of the most current

image encryption and decryption techniques. Int

J Adv Res Comput Sci 2019, 10(1): 9.

[27] Lagmiri SN, Elalami N and Elalami J. Color

and gray images encryption algorithm using

chaotic systems of different dimensions. Int J

Comput Sci Network Secur 2018, 18(1): 79–86.

[28] Parameshachari BD, Karappa R, Soyjaudah KS,

et al. Partial image encryption algorithm using

pixel position manipulation technique: the

smart copyback system. In: 2014 4th

international conference on artificial

intelligence with applications in engineering

and technology, Kota Kinabalu, Malaysia, 3–5

December 2014, pp.177–181. New York: IEEE.

[29] Stepan Bilan. Formation Methods, Models, and

Hardware Implementation of Pseudorandom

Number Generators: Emerging Research and

Opportunities. (2017). IGI Global, USA.- P.

301.

[30] Marco Tulio Ram´ırez Torres, Marcela Mej´ıa

Carlos, Jose S. Murgu ´ ´ıa Ibarra, Luis Javier

Ontanon Garcia. Partial Image Encryption

Using Cellular Automata. Computación y

Sistemas, Vol. 23, No. 4, 2019, pp. 1575–1582

[31] Ruslan Motornyuk, Andrii Bilan, and Stepan

Bilan. Research of Biometric Characteristics of

the Shape of the Ears Based on Multi-

Coordinate Methods. In book Biometric

Identifcation Technologies Based on Modern

Data Mining Methods. edited by Stepan Bilan,

Mohamed Elhoseny, D. Jude Hemanth pp. 177-

194.

[32] Stepan Bilan. Identification of rolling stock of

railways based on multi-projection image

processing methods. Proceedings of the 7th

International Conference "Information

Technology and Interactions" (IT&I-2020).

Workshops Proceedings Kyiv, Ukraine,

December 02-03, 2020. pp. 33-42.

[33] Ruslan Leonidovich Motornyuk, Stepan Bilan.

The Moving Object Detection and Research

Effects of Noise on Images Based on Cellular

Automata With a Hexagonal Coating Form and

Radon Transform. Handbook of Research on

Intelligent Data Processing and Information

Security Systems edited by Stepan Bilan,

Saleem Issa Al-Zoubi. pp. 330-359.

[34] Stepan Bilan, Mykola Bilan, Sergii Bilan.

Research of the method of pseudo-random

number generation based on asynchronous

cellular automata with several active cells.-

MATEC Web of Conferences, - Vol. 125,-

02018 (2017), - p. 1-6.

Contribution of Individual Authors to the

Creation of a Scientific Article (Ghostwriting

Policy)

Nashat Al Bdour has written, reviewed, and actively

participated in all the publication stages of this

manuscript.

Sources of Funding for Research Presented in a

Scientific Article or Scientific Article Itself

No funding was received for conducting this study.

Conflict of Interest

The author has no conflict of interest to declare that is

relevant to the content of this article.

Creative Commons Attribution License 4.0

(Attribution 4.0 International, CC BY 4.0)

This article is published under the terms of the

Creative Commons Attribution License 4.0

https://creativecommons.org/licenses/by/4.0/deed.en_

WSEAS TRANSACTIONS on INFORMATION SCIENCE and APPLICATIONS

DOI: 10.37394/23209.2023.20.13

Nashat Al Bdour

E-ISSN: 2224-3402

118

Volume 20, 2023