Trainable Regularization in Dense Image Matching Problems
This study examines the development of specialized models designed to solve image-matching problems. The purpose of this research is to develop a technique based on energy tensor aggregation for dense image matching. This task is relevant within the framework of computer systems since image comparis...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Ital Publication
2023-09-01
|
| Series: | HighTech and Innovation Journal |
| Subjects: | |
| Online Access: | https://hightechjournal.org/index.php/HIJ/article/view/437 |
| _version_ | 1797375510285647872 |
|---|---|
| author | Vladimir Zh. Kuklin Aslan A. Tatarkanov Alexander A. Umyskov |
| author_facet | Vladimir Zh. Kuklin Aslan A. Tatarkanov Alexander A. Umyskov |
| author_sort | Vladimir Zh. Kuklin |
| collection | DOAJ |
| description | This study examines the development of specialized models designed to solve image-matching problems. The purpose of this research is to develop a technique based on energy tensor aggregation for dense image matching. This task is relevant within the framework of computer systems since image comparison makes it possible to solve current problems such as reconstructing a three-dimensional model of an object, creating a panorama scene, ensuring object recognition, etc. This paper examines in detail the key features of the image matching process based on the use of binocular stereo reconstruction and the features of calculating energies during this process, and establishes the main parts of the proposed method in the form of diagrams and formulas. This research develops a machine learning model that provides solutions to image matching problems for real data using parallel programming tools. A detailed description of the architecture of the convolutional recurrent neural network that underlies this method is given. Appropriate computational experiments were conducted to compare the results obtained with the methods proposed in the scientific literature. The method discussed in this article is characterized by better efficiency, both in terms of the speed of work execution and the number of possible errors.
Doi: 10.28991/HIJ-2023-04-03-011
Full Text: PDF |
| first_indexed | 2024-03-08T19:25:28Z |
| format | Article |
| id | doaj.art-561e083b54d649f1930685a8276db349 |
| institution | Directory Open Access Journal |
| issn | 2723-9535 |
| language | English |
| last_indexed | 2024-03-08T19:25:28Z |
| publishDate | 2023-09-01 |
| publisher | Ital Publication |
| record_format | Article |
| series | HighTech and Innovation Journal |
| spelling | doaj.art-561e083b54d649f1930685a8276db3492023-12-26T09:29:16ZengItal PublicationHighTech and Innovation Journal2723-95352023-09-014361762910.28991/HIJ-2023-04-03-011143Trainable Regularization in Dense Image Matching ProblemsVladimir Zh. Kuklin0Aslan A. Tatarkanov1Alexander A. Umyskov2Institute of Design and Technology Informatics of RAS,Institute of Design and Technology Informatics of RAS,Institute of Design and Technology Informatics of RAS,This study examines the development of specialized models designed to solve image-matching problems. The purpose of this research is to develop a technique based on energy tensor aggregation for dense image matching. This task is relevant within the framework of computer systems since image comparison makes it possible to solve current problems such as reconstructing a three-dimensional model of an object, creating a panorama scene, ensuring object recognition, etc. This paper examines in detail the key features of the image matching process based on the use of binocular stereo reconstruction and the features of calculating energies during this process, and establishes the main parts of the proposed method in the form of diagrams and formulas. This research develops a machine learning model that provides solutions to image matching problems for real data using parallel programming tools. A detailed description of the architecture of the convolutional recurrent neural network that underlies this method is given. Appropriate computational experiments were conducted to compare the results obtained with the methods proposed in the scientific literature. The method discussed in this article is characterized by better efficiency, both in terms of the speed of work execution and the number of possible errors. Doi: 10.28991/HIJ-2023-04-03-011 Full Text: PDFhttps://hightechjournal.org/index.php/HIJ/article/view/437image matchingconvolutional recurrent neural networkstereo reconstructionmethod errorneural network architecture. |
| spellingShingle | Vladimir Zh. Kuklin Aslan A. Tatarkanov Alexander A. Umyskov Trainable Regularization in Dense Image Matching Problems HighTech and Innovation Journal image matching convolutional recurrent neural network stereo reconstruction method error neural network architecture. |
| title | Trainable Regularization in Dense Image Matching Problems |
| title_full | Trainable Regularization in Dense Image Matching Problems |
| title_fullStr | Trainable Regularization in Dense Image Matching Problems |
| title_full_unstemmed | Trainable Regularization in Dense Image Matching Problems |
| title_short | Trainable Regularization in Dense Image Matching Problems |
| title_sort | trainable regularization in dense image matching problems |
| topic | image matching convolutional recurrent neural network stereo reconstruction method error neural network architecture. |
| url | https://hightechjournal.org/index.php/HIJ/article/view/437 |
| work_keys_str_mv | AT vladimirzhkuklin trainableregularizationindenseimagematchingproblems AT aslanatatarkanov trainableregularizationindenseimagematchingproblems AT alexanderaumyskov trainableregularizationindenseimagematchingproblems |