Search

Your search keyword '"Do Thanh Ha"' showing total 31 results
Start Over Author "Do Thanh Ha"

Refine your results

more »

more »

more »

more »
31 results on '"Do Thanh Ha"'

1. 'NEIGHBOURHOOD FIRST POLICY' UNDER NARENDRA MODI: INDIA’S STRATEGIC CONCERNS IN SOUTH ASIA

Author

Do Thanh Ha

Subjects
Foreign Policy, Neighbourhood First, India, Strategy, Narendra Modi, Political theory, JC11-607, Law
Abstract

After becoming the Prime Minister of India, Narendra Modi made many efforts to consolidate his dominance in South Asia and expand India’s influence in the Indian Ocean. The “neighborhood first policy” was highlighted in India’s foreign policy. Based on historical and logical research methods and studies in international relations, the paper aims to examine the implementation of “neighborhood first policy” through bilateral relations between India and its neighboring countries. These relationships have been facing challenges from the more profound influence and interference of China as well as frequent terrorist incidents due to border disputes, which threatened regional security. Although this policy connected India with its neighboring South Asian countries, there were still challenges. It can be seen that the “neighborhood first policy” was Prime Minister Modi’s effort to nurture bilateral relationships, enhance regional connectivity, and strengthen and elevate India’s position in the region. The progress of the “Neighborhood First” policy is expected to continue more successfully in the next phase.

Published
2023

2. Hannotate: Flexible Annotation for Text Analytics from Anywhere

Author

To, Tan-Tai, Dao, Hoang, Nguyen, Huyen, Do, Thanh-Ha, Hoang, Tuan-Anh, Goos, Gerhard, Founding Editor, Hartmanis, Juris, Founding Editor, Bertino, Elisa, Editorial Board Member, Gao, Wen, Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Yung, Moti, Editorial Board Member, Pesquita, Catia, editor, Skaf-Molli, Hala, editor, Efthymiou, Vasilis, editor, Kirrane, Sabrina, editor, Ngonga, Axel, editor, Collarana, Diego, editor, Cerqueira, Renato, editor, Alam, Mehwish, editor, Trojahn, Cassia, editor, and Hertling, Sven, editor

Published
2023
Full Text
View/download PDF

4. Text Extraction Using Sparse Representation over Learning Dictionaries

Author

Do, Thanh-Ha, Nguyen, Thi Minh Huyen, Santosh, K. C., Barbosa, Simone Diniz Junqueira, Editorial Board Member, Filipe, Joaquim, Editorial Board Member, Ghosh, Ashish, Editorial Board Member, Kotenko, Igor, Editorial Board Member, Yuan, Junsong, Editorial Board Member, Zhou, Lizhu, Editorial Board Member, Santosh, K. C., editor, and Hegadi, Ravindra S., editor

Published
2019
Full Text
View/download PDF

5. Can We Understand Image Semantics from Conventional Neural Networks?

Author

Do, Thanh-Ha, Anh, Nguyen T. V., Dat, Nguyen T., Santosh, K. C., Barbosa, Simone Diniz Junqueira, Editorial Board Member, Filipe, Joaquim, Editorial Board Member, Ghosh, Ashish, Editorial Board Member, Kotenko, Igor, Editorial Board Member, Yuan, Junsong, Editorial Board Member, Zhou, Lizhu, Editorial Board Member, Santosh, K. C., editor, and Hegadi, Ravindra S., editor

Published
2019
Full Text
View/download PDF

16. A new approach for traffic-sign recognition using sparse representation over dictionary of local descriptors

Author

Le Ngoc Tuan, Nguyen Tien Dat, Nguyen Viet Thang, and Do Thanh Ha

Subjects
K-SVD, business.industry, Computer science, ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION, Cognitive neuroscience of visual object recognition, 020206 networking & telecommunications, Pattern recognition, 02 engineering and technology, Sparse approximation, Linear map, Robustness (computer science), 0202 electrical engineering, electronic engineering, information engineering, Traffic sign recognition, 020201 artificial intelligence & image processing, Artificial intelligence, business
Abstract

This paper is meant as the object recognition work done directly on images obtained from the autonomous car's camera. Normally, before the recognition stage, images need to process with the purpose of locating the candidate regions. This paper does not concentrate on the location stage, but focus on contributing a new descriptor for object recognition. Particularly, objects such as traffic-sign are recognized based on the sparse representation of object's descriptor over learned dictionary. The main idea is training the dictionary of local descriptors, or building the basis of a local descriptor space. Then, using the representation of description of object over this basis, we define a object's descriptor. We approve the robustness of the proposed descriptor under the linear transformation such as rotation or scale etc. Moreover, the stage of building the basis of descriptor space is done offline, thus computing time is quite reliably for the autonomous car. The experiments done on a database including traffic-sign images show that the proposed method is promising and competitive related to the state-of-the-art methods.

Published
2017

17. Spotting Symbol over Graphical Documents Via Sparsity in Visual Vocabulary

Author

Salvatore Tabbone, Do Thanh Ha, Oriol Ramos Terrades, Hanoi University of Science and Technology (HUST), Querying Graphics through Analysis and Recognition (QGAR), Department of Natural Language Processing & Knowledge Discovery (LORIA - NLPKD), Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS), Computer Vision Center (Centre de visio per computador) (CVC), and Universitat Autònoma de Barcelona (UAB)

Subjects
Vocabulary, Matching (graph theory), Property (programming), Computer science, business.industry, media_common.quotation_subject, [INFO.INFO-CV]Computer Science [cs]/Computer Vision and Pattern Recognition [cs.CV], 020206 networking & telecommunications, 02 engineering and technology, Spotting, computer.software_genre, Symbol (chemistry), [INFO.INFO-TT]Computer Science [cs]/Document and Text Processing, Computer Science::Computer Vision and Pattern Recognition, Similarity (psychology), 0202 electrical engineering, electronic engineering, information engineering, 020201 artificial intelligence & image processing, Shape context, Visual Word, Artificial intelligence, business, computer, ComputingMilieux_MISCELLANEOUS, Natural language processing, media_common
Abstract

This paper proposes a new approach to localize symbol in the graphical documents using sparse representations of local descriptors over learning dictionary. More specifically, a training database, being local descriptors extracted from documents, is used to build the learned dictionary. Then, the candidate regions into documents are defined following the similarity property between sparse representations of local descriptors. A vector model for candidate regions and for a query symbol is constructed based on the sparsity in a visual vocabulary where the visual words are columns in the learned dictionary. The matching process is performed by comparing the similarity between vector models. The first evaluation on SESYD database demonstrates that the proposed method is promising.

Published
2017

20. Noise suppression over bi-level graphical documents by sparse representation

Author

Do, Thanh Ha, Tabbone, Salvatore, Ramos-Terrades, Oriol, Do, Thanh Ha, Querying Graphics through Analysis and Recognition (QGAR), Department of Natural Language Processing & Knowledge Discovery (LORIA - NLPKD), Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria), CVC, Computer Vision Center (Centre de visio per computador) (CVC), Universitat Autònoma de Barcelona (UAB)-Universitat Autònoma de Barcelona (UAB), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), and Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)

Subjects
K-SVD, [INFO.INFO-TI] Computer Science [cs]/Image Processing [eess.IV], [INFO.INFO-TI]Computer Science [cs]/Image Processing [eess.IV], noise removal, graphical documents, Sparse representation, learned dictionary
Abstract

In this paper, we explore the use of learning algorithm (K-SVD) for building dictionaries adapted to the image properties. In addition, in our model, we also modeled the energy of the noise basing on the function of the normalized cross-correlation between noised and non noised documents identified in training set. We have evaluated this method on the Grec2005 dataset. The experimental results demonstrate the robustness of our approach by comparing it with state-of-the-art methods., Dans cet article, nous explorons l'utilisation de l'algorithme d'apprentissage (K-SVD) pour construire des dictionnaires adaptés aux documents graphiques. En plus, dans notre modèle, nous avons également modélisé l'énergie du bruit à partir de la fonction de la corrélation croisée normalisée entre les documents bruités et non bruités définis dans notre base d'apprentissage. Nous avons évalué cette méthode sur la base de données Grec2005. Les résultats expérimentaux démontrent la robustesse de notre approche en comparant à des méthodes de l'état de l'art.

Published
2012

21. Présentations parcimonieuses sur dictionnaire d'apprentissage pour l'analyse de documents

Author

Do, Thanh Ha, UL, Thèses, Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS), Université de Lorraine, Salvatore Tabbone, Oriol Ramos Terrades, and Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)

Subjects
Dictionnaire d'apprentissage, Traitement d'images -- Techniques numériques, Separation text/graphic, Symbol spotting, [INFO.INFO-OH]Computer Science [cs]/Other [cs.OH], Réduction du bruit, Localisation de symboles, Symbol recognition, Reconnaissance de symboles, Learning algorithms, Représentations parcimonieuses, Sparse representations, Mots visuels, Systèmes experts (informatique), [INFO.INFO-OH] Computer Science [cs]/Other [cs.OH], Reconnaissance des formes (informatique), Visual words, Algorithme apprentissage, Learned dictionary, Séparation texte/graphique, Removal noise, Reconnaissance optique des caractères
Abstract

In this thesis, we focus on how sparse representations can help to increase the performance of noise removal, text region extraction, pattern recognition and spotting symbols in graphical documents. To do that, first of all, we give a survey of sparse representations and its applications in image processing. Then, we present the motivation of building learning dictionary and efficient algorithms for constructing a learning dictionary. After describing the general idea of sparse representations and learned dictionary, we bring some contributions in the field of symbol recognition and document processing that achieve better performances compared to the state-of-the-art. These contributions begin by finding the answers to the following questions. The first question is how we can remove the noise of a document when we have no assumptions about the model of noise found in these images? The second question is how sparse representations over learned dictionary can separate the text/graphic parts in the graphical document? The third question is how we can apply the sparse representation for symbol recognition? We complete this thesis by proposing an approach of spotting symbols that use sparse representations for the coding of a visual vocabulary, Dans cette thèse, nous nous concentrons sur comment les représentations parcimonieuses peuvent aider à augmenter les performances pour réduire le bruit, extraire des régions de texte, reconnaissance des formes et localiser des symboles dans des documents graphiques. Pour ce faire, tout d'abord, nous donnons une synthèse des représentations parcimonieuses et ses applications en traitement d'images. Ensuite, nous présentons notre motivation pour l'utilisation de dictionnaires d'apprentissage avec des algorithmes efficaces pour les construire. Après avoir décrit l'idée générale des représentations parcimonieuses et du dictionnaire d'apprentissage, nous présentons nos contributions dans le domaine de la reconnaissance de symboles et du traitement des documents en les comparants aux travaux de l'état de l'art. Ces contributions s'emploient à répondre aux questions suivantes: La première question est comment nous pouvons supprimer le bruit des images où il n'existe aucune hypothèse sur le modèle de bruit sous-jacent à ces images ? La deuxième question est comment les représentations parcimonieuses sur le dictionnaire d'apprentissage peuvent être adaptées pour séparer le texte du graphique dans des documents? La troisième question est comment nous pouvons appliquer la représentation parcimonieuse à reconnaissance de symboles? Nous complétons cette thèse en proposant une approche de localisation de symboles dans les documents graphiques qui utilise les représentations parcimonieuses pour coder un vocabulaire visuel

Published
2014

22. Vietnam and China: ideological bedfellows, strange dreamers

Author

Do Thanh Hai

Subjects
vietnam, china, ideological ties, economic interdependence, south china sea, identity, Political institutions and public administration - Asia (Asian studies only), JQ1-6651, Social sciences and state - Asia (Asian studies only), H53
Abstract

The article examines the evolution of Vietnam–China relations in the post-Cold War to identify the factors that shapes its dynamics. It finds that the bilateral relations have been significantly changed overtime, from strategic antagonism to ideology-shared partnership and subsequently from economic sistership to security rivalry. The two countries’ worldviews have gradually diverged. Given geographical nearness is a constant, such a course is driven by the interactions of four factours, including shifts in security environment, internal factional politics, economic calculations, and rising nationalism. At different points of time, some factor rises dominantly and these others matter less. However, it is observed that since the normalization in 1991 shared ideology figured less and less significantly.

Published
2021
Full Text
View/download PDF

23. Document Noise Removal using Sparse Representations over Learned Dictionary

Author

Do Thanh-Ha, Salvatore Tabbone, Oriol Ramos Terrades, Querying Graphics through Analysis and Recognition (QGAR), Department of Natural Language Processing & Knowledge Discovery (LORIA - NLPKD), Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria), Computer Vision Center (Centre de visio per computador) (CVC), Universitat Autònoma de Barcelona (UAB), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), and Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)

Subjects
K-SVD, Training set, Cross-correlation, Noise energy, business.industry, Computer science, Speech recognition, Noise reduction, Computer Science::Information Retrieval, 020206 networking & telecommunications, Pattern recognition, 02 engineering and technology, Sparse approximation, Robustness (computer science), [INFO.INFO-TI]Computer Science [cs]/Image Processing [eess.IV], 0202 electrical engineering, electronic engineering, information engineering, 020201 artificial intelligence & image processing, Artificial intelligence, business, Noise removal
Abstract

International audience; In this paper, we propose an algorithm for denoising document images using sparse representations. Following a training set, this algorithm is able to learn the main document characteristics and also, the kind of noise included into the documents. In this perspective, we propose to model the noise energy based on the normalized cross-correlation between pairs of noisy and non-noisy documents. Experimental results on several datasets demonstrate the robustness of our method compared with the state-of-the-art.

Published
2013

24. Arabidopsis thaliana Raf22 protein kinase maintains growth capacity during postgerminative growth arrest under stress.

Author

Hwang, Jae‐Ung, Yim, Sojeong, Do, Thanh Ha Thi, Kang, Joohyun, and Lee, Youngsook

Subjects
PLANT growth, GERMINATION, PLANT embryology, PROTEIN kinases, ABSCISIC acid, ARABIDOPSIS thaliana, GENETIC overexpression
Abstract

Abstract: When seeds are exposed to drought and salinity during germination, newly germinated embryos stop growth and enter a quiescent state called postgerminative growth (PGG) arrest. PGG arrest protects embryos from the stress, but it is not known how PGG is restored from the arrest when stress is eased. In this study, we show that under stress‐ or abscisic acid‐induced PGG arrest conditions, Arabidopsis thaliana Raf‐type protein kinase 22 (AtRaf22) overexpression accelerated photoautotrophic seedling establishment, whereas atraf22 knockout mutations enhanced the arrest. Furthermore, when the stress intensity was reduced subsequently, AtRaf22 overexpression plants resumed growth and accomplished photoautotrophic transition much faster than the knockout or wild‐type plants. These results suggest that AtRaf22 activity is important for maintaining growth capacity during stress‐induced PGG arrest, which is most likely critical for competitive growth when the stress subsides and plants resume growth. Such a factor has not been reported before. [ABSTRACT FROM AUTHOR]

Published
2018
Full Text
View/download PDF

27. Text/graphic separation using a sparse representation with multi-learned dictionaries.

Author

Do, Thanh-Ha, Tabbone, Salvatore, and Ramos-Terrades, Oriol

Abstract

In this paper, we propose a new approach to extract text regions from graphical documents. In our method, we first empirically construct two sequences of learned dictionaries for the text and graphical parts respectively. Then, we compute the sparse representations of all different sizes and non-overlapped document patches in these learned dictionaries. Based on these representations, each patch can be classified into the text or graphic category by comparing its reconstruction errors. Same-sized patches in one category are then merged together to define the corresponding text or graphic layers which are combined to create a final text/graphic layer. Finally, in a post-processing step, text regions are further filtered out by using some learned thresholds. [ABSTRACT FROM PUBLISHER]

Published
2012

28. Solution on reducing different settlement on road embankment and abutment pier in Long Sonbrigde, Long An province

Author

Đỗ Thanh Hải and Nguyễn Trung Hậu

Subjects
trụ xi măng đất, tỉ lệ trộn, chênh lệch lún, Biotechnology, TP248.13-248.65
Abstract

Different settlement between road embankment on the soft soil and the bridge is very popular researched area. This paper experimented on the mixture of soil and cement with the different ratio of cement/soil (X/D): 150kg/m3, 200kg/m3, 250kg/m3and cured in the natural soil put on the box. The result applied for the road embankment to Long Son –Long An. The Plaxis 2D to model soil-cement coloumn with diameter 600mm and 6m length and settlement of abutment pier is used in this problem. After analyzed, author suggest the ratio of cement/soil is 200kg/m3to improve the soft soil under the road embankment. It is shown that the different settlement now is reducing from 17.344 cm to 6.685 cm (reducing about 61.5%).

Published
2016

29. Effect of leaching on shear strength of marine clay in Can Gio

Author

Do Thanh Hai, Huynh Tan Phat, Vo Duy Phuoc, and Le Hoang Phuong

Subjects
salinity, cohesion, internal friction angle, Biotechnology, TP248.13-248.65
Abstract

Effect of leaching on shear strength of marine clay at Can Gio, Ho Chi Minh City is investigated in laboratory tests. Soil specimens were collected from three sites with different soil salinity. Testing includes shear strength and initial soil salinity. Leaching process is performed by modified equipment in laboratory with fresh water, then specimen is subjected to direct shear test to determine shear strength of leached and unleached soil. It is indicated that leaching affects the shear strength as reducing cohesion value (c, kPa) from 30% – 70% and internal friction angle (φ,o) from 8% - 20%.

Published
2012

30. New Approach for Symbol Recognition Combining Shape Context of Interest Points with Sparse Representation

Author

Salvatore Tabbone, Oriol Ramos Terrades, Thanh Ha Do, Do, Thanh Ha, Querying Graphics through Analysis and Recognition (QGAR), Department of Natural Language Processing & Knowledge Discovery (LORIA - NLPKD), Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria), Computer Vision Center (Centre de visio per computador) (CVC), Universitat Autònoma de Barcelona (UAB), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), and Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)

Subjects
Vocabulary, Similarity (geometry), Computer science, business.industry, media_common.quotation_subject, 020206 networking & telecommunications, Pattern recognition, 02 engineering and technology, Sparse approximation, computer.software_genre, Symbol (chemistry), Ranking (information retrieval), Ranking, [INFO.INFO-TI] Computer Science [cs]/Image Processing [eess.IV], [INFO.INFO-TI]Computer Science [cs]/Image Processing [eess.IV], 0202 electrical engineering, electronic engineering, information engineering, Benchmark (computing), 020201 artificial intelligence & image processing, Shape context, Visual Word, Artificial intelligence, business, computer, Natural language processing, media_common
Abstract

International audience; In this paper, we propose a new approach for symbol description. Our method is built based on the combination of shape context of interest points descriptor and sparse representation. More specifically, we first learn a dictionary describing shape context of interest point descriptors. Then, based on information retrieval techniques, we build a vector model for each symbol based on its sparse representation in a visual vocabulary whose visual words are columns in the learned dictionary. The retrieval task is performed by ranking symbols based on similarity between vector models. The evaluation of our method, using benchmark datasets, demonstrates the validity of our approach and shows that it outperforms related state-of-theart methods.

Published
2013

31. Text/graphic separation using a sparse representation with multi-learned dictionaries

Author

Do, T. -H, Tabbone, S., Oriol Ramos Terrades, Querying Graphics through Analysis and Recognition (QGAR), Department of Natural Language Processing & Knowledge Discovery (LORIA - NLPKD), Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS), Computer Vision Center (Centre de visio per computador) (CVC), Universitat Autònoma de Barcelona (UAB), Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria), and Do, Thanh Ha

Subjects
[INFO.INFO-TI] Computer Science [cs]/Image Processing [eess.IV], [INFO.INFO-TI]Computer Science [cs]/Image Processing [eess.IV], Document Understanding, Graphics Recognition, Layout Analysis, ComputingMethodologies_COMPUTERGRAPHICS
Abstract

International audience; In this paper, we propose a new approach to extract text regions from graphical documents. In our method, we first empirically construct two sequences of learned dictionaries for the text and graphical parts respectively. Then, we compute the sparse representations of all different sizes and non-overlapped document patches in these learned dictionaries. Based on these representations, each patch can be classified into the text or graphic category by comparing its reconstruction errors. Same-sized patches in one category are then merged together to define the corresponding text or graphic layers which are combined to createfinal text/graphic layer. Finally, in a post-processing step, text regions are further filtered out by using some learned thresholds.

Catalog

Books, media, physical & digital resources
See catalog results