Intellectualization of Moving Transport Objects Recognition
The purpose of the article is to research, analyze, and consider current problems and prospects for the development of software for the recognition of transport objects based on the use of pattern recognition theory, methods, and tools of artificial intelligence, and different types of neural netwo...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
KNUKiM Publishing Centre
2022-12-01
|
| Series: | Цифрова платформа: інформаційні технології в соціокультурній сфері |
| Subjects: | |
| Online Access: | http://infotech-soccult.knukim.edu.ua/article/view/270147 |
| _version_ | 1828062145468694528 |
|---|---|
| author | Ольга Ткаченко Дмитро Бабиченко |
| author_facet | Ольга Ткаченко Дмитро Бабиченко |
| author_sort | Ольга Ткаченко |
| collection | DOAJ |
| description |
The purpose of the article is to research, analyze, and consider current problems and prospects for the development of software for the recognition of transport objects based on the use of pattern recognition theory, methods, and tools of artificial intelligence, and different types of neural networks.
The research methodology is basic methods and algorithms of pattern recognition, methods and means of artificial intelligence, and different types of neural networks. The article considers the main problems of intellectualization of processes occurring in transport. The main attention is paid to the intellectualization of the processes of transport objects’ recognition. The article analyzes the most common recognition methods. A study of these methods and approaches to the recognition of moving vehicles is conducted. It is determined which methods have high and which have low computational complexity. Among the considered methods are those that recognize static transport objects (primarily transport infrastructure objects) using intelligent technologies, statistical, probabilistic, and other methods. The main attention is paid to methods that recognize dynamic transport objects. The basis of intellectualization of the processes of recognition of this group of objects is the use of neural networks, in particular convolutional, recurrent, neural networks with long short-term memory (LSTM), etc.
The scientific novelty of the research is the analysis of modern methods of moving transport objects’ recognition, the results of which can be used in the development of their software product. The article emphasizes that the proposed modern approach to the recognition of transport objects (moving vehicles and transport infrastructure) involves solving a wide range of problems based on the use of intelligent technologies, in particular neural networks.
Conclusions. The most common methods for solving current problems of recognition of transport objects (vehicles and transport infrastructure) have been investigated and analyzed. Based on the analysis of methods for recognizing moving vehicles, it has been determined which methods have high and low computational complexity. The basis of the intellectualization of the recognition processes for this group of transport objects is the use of various neural networks.
|
| first_indexed | 2024-04-10T22:24:39Z |
| format | Article |
| id | doaj.art-e256a998e036411aac129486436f4c5d |
| institution | Directory Open Access Journal |
| issn | 2617-796X 2618-0049 |
| language | English |
| last_indexed | 2024-04-10T22:24:39Z |
| publishDate | 2022-12-01 |
| publisher | KNUKiM Publishing Centre |
| record_format | Article |
| series | Цифрова платформа: інформаційні технології в соціокультурній сфері |
| spelling | doaj.art-e256a998e036411aac129486436f4c5d2023-01-17T11:07:03ZengKNUKiM Publishing CentreЦифрова платформа: інформаційні технології в соціокультурній сфері2617-796X2618-00492022-12-015210.31866/2617-796X.5.2.2022.270147Intellectualization of Moving Transport Objects RecognitionОльга Ткаченко0Дмитро Бабиченко1State University of Infrastructure and TechnologyState University of Infrastructure and Technology The purpose of the article is to research, analyze, and consider current problems and prospects for the development of software for the recognition of transport objects based on the use of pattern recognition theory, methods, and tools of artificial intelligence, and different types of neural networks. The research methodology is basic methods and algorithms of pattern recognition, methods and means of artificial intelligence, and different types of neural networks. The article considers the main problems of intellectualization of processes occurring in transport. The main attention is paid to the intellectualization of the processes of transport objects’ recognition. The article analyzes the most common recognition methods. A study of these methods and approaches to the recognition of moving vehicles is conducted. It is determined which methods have high and which have low computational complexity. Among the considered methods are those that recognize static transport objects (primarily transport infrastructure objects) using intelligent technologies, statistical, probabilistic, and other methods. The main attention is paid to methods that recognize dynamic transport objects. The basis of intellectualization of the processes of recognition of this group of objects is the use of neural networks, in particular convolutional, recurrent, neural networks with long short-term memory (LSTM), etc. The scientific novelty of the research is the analysis of modern methods of moving transport objects’ recognition, the results of which can be used in the development of their software product. The article emphasizes that the proposed modern approach to the recognition of transport objects (moving vehicles and transport infrastructure) involves solving a wide range of problems based on the use of intelligent technologies, in particular neural networks. Conclusions. The most common methods for solving current problems of recognition of transport objects (vehicles and transport infrastructure) have been investigated and analyzed. Based on the analysis of methods for recognizing moving vehicles, it has been determined which methods have high and low computational complexity. The basis of the intellectualization of the recognition processes for this group of transport objects is the use of various neural networks. http://infotech-soccult.knukim.edu.ua/article/view/270147transport objectsintelligent technologiesintellectualizationneural networksrecognition methodsrecognition of transport objects |
| spellingShingle | Ольга Ткаченко Дмитро Бабиченко Intellectualization of Moving Transport Objects Recognition Цифрова платформа: інформаційні технології в соціокультурній сфері transport objects intelligent technologies intellectualization neural networks recognition methods recognition of transport objects |
| title | Intellectualization of Moving Transport Objects Recognition |
| title_full | Intellectualization of Moving Transport Objects Recognition |
| title_fullStr | Intellectualization of Moving Transport Objects Recognition |
| title_full_unstemmed | Intellectualization of Moving Transport Objects Recognition |
| title_short | Intellectualization of Moving Transport Objects Recognition |
| title_sort | intellectualization of moving transport objects recognition |
| topic | transport objects intelligent technologies intellectualization neural networks recognition methods recognition of transport objects |
| url | http://infotech-soccult.knukim.edu.ua/article/view/270147 |
| work_keys_str_mv | AT olʹgatkačenko intellectualizationofmovingtransportobjectsrecognition AT dmitrobabičenko intellectualizationofmovingtransportobjectsrecognition |