ITM Web of Conferences 59, 04009 (2024) https://doi.org/10.

1051/itmconf/20245904009
HMMOCS-II 2023

Artificial intelligence's ability to process images

Islam A. Magomedov1*, Amina Sh. Khafizova2, and Fardiana R. Ketova3
1
Kadyrov Chechen State University, Grozny, Russia
2
Kazan National Research Technical University named after A. N. Tupolev-KAI, Kazan, Russia
3Kabardino-Balkarian State University, Nalchik, Russia

Abstract. Penetration of technology into different fields and their dominion

over them after a while is undoubtable. New technologies lure our souls and
the dependency on them is inextricable once having tasted the fruit. We as a
human easily addict to the convenient and comforting songs. As well as
when they do our work for us in a more preferable and presentable way. This
work was done to illustrate the artificial intelligence use in image
processing. Its capabilities and issues are the key factors of the concerns of
the work. Therefore, the work outlines the basic ideas behind artificial
intelligence and its use in different areas and making accent on the
capabilities of the technology when using tools powered by artificial
intelligence to understand the meaning of the provided picture. Possible
solutions are also described.

1 Introduction
As we delve into the realm of Artificial intelligence, we can see its vast and endless
capabilities in different fields. Artificial intelligence is a modern revolution that transforms
our well-established routine lives into an unimaginable futuristic world. The technology
brings a simple idea. The idea that human capabilities might be mimicked and even outrun.
The very idea was fantasy a few decades ago, only seen in movie. With the technological
progress, occurrence of different methods and improvement of algorithms it was possible for
the technology to emerge in the state it is now.
Artificial intelligence will defy the nearest future. Its capabilities allow it to penetrate
almost to any field and if correctly used to level them up. It brings many possibilities. The
vital fields are using artificial intelligence to the state that it is almost impossible for a
common man to notice its application. It is so well embedded, and the realization of its use
in some cases puts some people to unbelief. It is true that its use is spreading like well-built
viruses, finding its victims and sending them to the shocking level. However, in this case
meaning the positive impact.
The importance of artificial intelligence is dictated by its capabilities and its application.
Therefore, the clear knowledge of the technology must be a well-known for its best
application. The technology brings two different yet with the mirroring abilities sides. There
are clear positive and negative sides. There is also mixed effect as well. From one angle, the
use of technology can satisfy the user (any industry) and negatively effect, for example, the

* Corresponding author: [email protected]

© The Authors, published by EDP Sciences. This is an open access article distributed under the terms of the Creative Commons
Attribution License 4.0 (https://creativecommons.org/licenses/by/4.0/).
ITM Web of Conferences 59, 04009 (2024) https://doi.org/10.1051/itmconf/20245904009
HMMOCS-II 2023

employee who works there. At the end, it does indicate some similar tendency as any other
technology, but with the slightest hints [1, 2].
Its usability in different fields brings elevation of the system. The most notable changes
it provided to the entertainment industry. It immediately locked itself in the core of tools used
in the field. bringing them to the shocking state. These tools are capable of generating almost
anything from text to videos. Of course, there are still some areas that lack quality and
correctness. In general, its capabilities, in short time, transformed entertainment industry that
backlashed negativity. Nonetheless, the following article will be focusing on studying the
capabilities of artificial intelligence to process and understand the images.

2 The science behind processing images through artificial

intelligence
The first step in this journey is to have an image ready, which can be taken through any
convenient device that is capable of generating appropriate formats for the further analysis.
The image selected then can be directly passed through different tools to enrich some parts
and to eliminate some issues such as nose, blur, glitches and so on. The other case is to
analyses in its raw state without preprocessing. Computer vision is used to extract patterns
and points of different object that it will use lately to understand the picture.
The next step is to utilize machine learning through various ways. It can be done by three
methods: supervised, unsupervised and simi-supervised learning. Through these methods the
algorithms learn to map input data through of variation of labeling processes. Deep learning
applying convolutional neural networks understands tasks starting from the small details and
gradually figuring out more complex features in the image. Meaning, it will start from the
scratch and move further up by learning piece to piece. Neural network architectures consist
of convolutional layers, pooling layers, and activation functions, and can be fine-tuned using
transfer learning. Transfer learning can be described as a feature that recalls what it studied
earlier and, on that recall, constructs further improvements. The other step is to cut picture
into segments (called segmentation process) to understand and analysis it more detail. There
are many different steps and technologies used to analyse the image, but in more general way
it is as explained above [3, 4].

3 Application of image processing

One thing is the capabilities of understanding the information on the images, and another its
possible applications. Its application is vast and in demand. There are many fields that can
benefit from this technology. Therefore, the following fields embed the technology to
thoroughly examine the imagery: medicine, entertainment and agriculture.

3.1 Medicine
As we all know, there are many areas in medicine involved with image analysis. The image
analyzing technology can be utilized for the processing of X-rays, MRIs, CT scans. The
technology can be used to diagnose and monitor diseases. The enrichment of the technology
allows its users to go more deeper in a sense of its provision of wider scope of results that
can be collected. With the improvement on diagnosis, it can increase the correctness of the
outcomes drastically [5].

2
ITM Web of Conferences 59, 04009 (2024) https://doi.org/10.1051/itmconf/20245904009
HMMOCS-II 2023

3.2 Entertainment

Entertainment can be named as a field that introduced the capabilities of neural network to
the masses. This is also the one that utilizes the technology the most compared to other fields.
Application of the technology in this field is dramatic and diverse. The technology can be
found in many different apps and software. It provides much more easiness to the users and
flowless experience. Examples of such implementation can be found in almost every device
that has cameras or software to process images. For instance, the features of the face is refined
in real time by understanding and analyzing the image that been fed. Most of the time it is a
frontal camera that utilizes such enrichment. Augmented and virtual realities also use image
processing in real time to construct and augment the outcomes. There are many various
applications, but will not be covered in this work [6, 7].

3.3 Agriculture
Agriculture is the sector that is thriving on using new technologies. Its first task is to
withstand the pressure from the overpopulation and global warming. These issues put the
sector under harsh environments that needs to be addressed before it gets worse. The real
solution is utilization of new technologies and in this case the use of artificial intelligence.
The use of artificial intelligence pushes this sector to the new level, of course with the
combination of other technology. The use of image processing is also a pivotal part in
agricultural sector. It enables the system to monitor and analyze the overall process without
involvement of the workforce. It can be compared to mechanism that are powered by sight,
that allows them to function without outer control. The technology will only improve and
further transform the agricultural sector in the positive direction [8, 9].

4 Examples of the use

To understand how well the technology works, some analysis was done. DALL.E was used
as an example. Three simple but related images were provided (Figure 1). The simple
prompts were given. First one was about description. The tool described all three images
correctly. By observing Figure 2 one can see the description. Then the tool was asked to more
closely describe the images to find correlations. Overall, the outcomes were clear, except two
detail. The tool suggested that the guy in the adidas t-short occurs in two images, which is in
reality wrong. The second one was about the department where the event took place. The tool
suggested that it probably will be geography department. The idea for that outcome might be
the result of a map on the second image that conflicted the tool to think it has relation to the
suggested place.

Fig. 1. provided images to the DALL.E.

3
ITM Web of Conferences 59, 04009 (2024) https://doi.org/10.1051/itmconf/20245904009
HMMOCS-II 2023

Fig. 2. The provided description to the image.

5 Future perspectives
Looking ahead, the future of image processing holds promising developments in diverse
areas. In the medical field, there's a trajectory toward more advanced imaging techniques like
3D and functional imaging, coupled with artificial intelligence for improved diagnostics.
Augmented and virtual reality experiences are set to become even more realistic and
immersive, impacting education, training, and entertainment. Autonomous vehicles are
expected to benefit from ongoing image processing advancements, enhancing object
recognition and decision-making for safer transportation. Human-computer interaction will
likely see improvements in gesture recognition and emotion analysis, influencing
applications in gaming, healthcare, and smart environments. Image processing's role in
environmental monitoring, particularly through satellite imagery analysis, will continue to
aid resource management, disaster response, and climate studies. In robotics and automation,
sophisticated systems capable of navigating complex environments and performing intricate
tasks are anticipated. The evolution of smart cities will rely on image processing for traffic
management, public safety, and urban planning. Security systems are likely to advance with
facial recognition and behavior analysis. Personalized user experiences, quantum image
processing, computational photography innovations, and the potential integration of image
processing in brain-computer interfaces also mark exciting avenues for exploration in the
future. These trends underscore the ongoing integration of image processing into various
aspects of our lives, driven by technological advancements and evolving applications [9, 10,
11,12].

6 Conclusion
To conclude, the following work was done to show the abilities of artificial intelligence when
dealing with images. Its capabilities over analysis and understanding was the main focus.
Artificial intelligence use in modern time is vast, and its wide application is due to its
capabilities to be implemented almost everywhere. Similarly, the technology is capable of

4
ITM Web of Conferences 59, 04009 (2024) https://doi.org/10.1051/itmconf/20245904009
HMMOCS-II 2023

distinguishing the different parts of the images. Thus, the technology is uses in differed fields
and areas involved with the image processing method. Some of which were covered in this
work, such is its use in medicine, agriculture, and entertainment. The technology is a
promising technology not just in image processing field, but in many others. Some possible
outcomes were also discussed in this work.

References
1. I. Magomedov, E. Belashova, M.-D. Bersanov, E3S Web of Conf. 402, 03042 (2023)
2. V.A. Gerasimov, M.G. Nuriev, D.A. Gashigullin, International Russian Automation
Conference (2022)
3. S. Hussain, P. Dixit, S. Hussain, Image Processing in Artificial Intelligence,
International Journal of Scientific Research in Computer Science Engineering and
Information Technology (2020)
4. J. Long, E. Shelhamer and T. Darrell, Fully convolutional networks for semantic
segmentation, 2015 IEEE Conference on Computer Vision and Pattern Recognition
(CVPR), Boston, MA, USA, 2015, pp. 3431-3440 (2015)
5. R. Obuchowicz, M. Strzelecki, A. Piorkowski, Artificial Intelligence in Medical Imaging
and Image Processing, MDPI (2023)
6. A. Singh, A Survey of AI Text-to-Image and AI Text-to-Video Generators, Conference:
2023 4th International Conference on Artificial Intelligence, Robotics and Control
(AIRC) (2023)
7. R. Songja, I. Promboot, B. Haetanurak, C. Kerdvibulvech, Deepfake AI images: should
deepfakes be banned in Thailand? AI Ethics (2023)
8. I. Magomedov, M.M. Khulamkhanova, N.A. Staroverova, BIO Web Conf. 63, 05010
(2023)
9. I.A. Magomedov et al., IOP Conf. Ser.: Earth Environ. Sci. 548, 032029 (2020)
10. J. Shabbir, T. Anwer, Artificial Intelligence and its Role in Near Future. ArXiv. 14(8),
1-11 (2015)
11. M.I. Jordan, T.M. Mitchell, Science 349(6245), 255-260 (2023)
12. S.N. Cherny, R.F. Gibadullin, The Recognition of Handwritten Digits Using Neural
Network Technology, International Conference on Industrial Engineering, Applications
and Manufacturing (ICIEAM), Sochi, Russian (2022)