Development of software for the segmentation of text areas in real-scene images
This article discusses the design and development of a neural network algorithm for the segmentation of text areas in real-scene images. After reviewing the available neural network models, the U-net model was chosen as a basis. Then an algorithm for detecting text areas in real-scene images was pro...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Samara National Research University
2022-10-01
|
| Series: | Компьютерная оптика |
| Subjects: | |
| Online Access: | https://computeroptics.ru/eng/KO/Annot/KO46-5/460513e.html |
| _version_ | 1797649949772480512 |
|---|---|
| author | V.A. Lobanova Yu.A. Ivanova |
| author_facet | V.A. Lobanova Yu.A. Ivanova |
| author_sort | V.A. Lobanova |
| collection | DOAJ |
| description | This article discusses the design and development of a neural network algorithm for the segmentation of text areas in real-scene images. After reviewing the available neural network models, the U-net model was chosen as a basis. Then an algorithm for detecting text areas in real-scene images was proposed and implemented. The experimental training of the network allows one to define the neural network parameters such as the size of input images and the number and types of the network layers. Bilateral and low-pass filters were considered as a preprocessing stage. The number of images in the KAIST Scene Text Database was increased by applying rotations, compression, and splitting of the images. The results obtained were found to surpass competing methods in terms of the F-measure value. |
| first_indexed | 2024-03-11T15:54:29Z |
| format | Article |
| id | doaj.art-f1ce227202f14fd08f42c7673ffd883a |
| institution | Directory Open Access Journal |
| issn | 0134-2452 2412-6179 |
| language | English |
| last_indexed | 2024-03-11T15:54:29Z |
| publishDate | 2022-10-01 |
| publisher | Samara National Research University |
| record_format | Article |
| series | Компьютерная оптика |
| spelling | doaj.art-f1ce227202f14fd08f42c7673ffd883a2023-10-25T11:47:47ZengSamara National Research UniversityКомпьютерная оптика0134-24522412-61792022-10-0146579080010.18287/2412-6179-CO-1047Development of software for the segmentation of text areas in real-scene imagesV.A. Lobanova0Yu.A. Ivanova1Tomsk Polytechnic UniversityTomsk Polytechnic UniversityThis article discusses the design and development of a neural network algorithm for the segmentation of text areas in real-scene images. After reviewing the available neural network models, the U-net model was chosen as a basis. Then an algorithm for detecting text areas in real-scene images was proposed and implemented. The experimental training of the network allows one to define the neural network parameters such as the size of input images and the number and types of the network layers. Bilateral and low-pass filters were considered as a preprocessing stage. The number of images in the KAIST Scene Text Database was increased by applying rotations, compression, and splitting of the images. The results obtained were found to surpass competing methods in terms of the F-measure value.https://computeroptics.ru/eng/KO/Annot/KO46-5/460513e.htmldeep learningu-net architectureimage processingimage segmentationtext areasreal scenes images |
| spellingShingle | V.A. Lobanova Yu.A. Ivanova Development of software for the segmentation of text areas in real-scene images Компьютерная оптика deep learning u-net architecture image processing image segmentation text areas real scenes images |
| title | Development of software for the segmentation of text areas in real-scene images |
| title_full | Development of software for the segmentation of text areas in real-scene images |
| title_fullStr | Development of software for the segmentation of text areas in real-scene images |
| title_full_unstemmed | Development of software for the segmentation of text areas in real-scene images |
| title_short | Development of software for the segmentation of text areas in real-scene images |
| title_sort | development of software for the segmentation of text areas in real scene images |
| topic | deep learning u-net architecture image processing image segmentation text areas real scenes images |
| url | https://computeroptics.ru/eng/KO/Annot/KO46-5/460513e.html |
| work_keys_str_mv | AT valobanova developmentofsoftwareforthesegmentationoftextareasinrealsceneimages AT yuaivanova developmentofsoftwareforthesegmentationoftextareasinrealsceneimages |