On the possibility of structure learning-based scene character detector

Yugo Terada, Rong Huang, Yaokai Feng, Seiichi Uchida

Research output: Contribution to journalConference article

4 Citations (Scopus)

Abstract

In this paper, we propose a structure learning-based scene character detector which is inspired by the observation that characters have their own inherent structures compared with the background. Graphs are extracted from the thinned binary image to represent the topological line structures of scene contents. Then, a graph classifier, namely gBoost classifier, is trained with the intent to seek out the inherent structures of character and the counterparts of non-character. The experimental results show that the proposed detector achieves the remarkable classification performance with the accuracy of about 70%, which demonstrates the existence and separability of the inherent structures.

Original languageEnglish
Article number6628666
Pages (from-to)472-476
Number of pages5
JournalProceedings of the International Conference on Document Analysis and Recognition, ICDAR
DOIs
Publication statusPublished - Dec 11 2013
Event12th International Conference on Document Analysis and Recognition, ICDAR 2013 - Washington, DC, United States
Duration: Aug 25 2013Aug 28 2013

Fingerprint

Classifiers
Detectors
Binary images

All Science Journal Classification (ASJC) codes

  • Computer Vision and Pattern Recognition

Cite this

@article{6e8f95393f5d433c9187f445e3c1ce50,
title = "On the possibility of structure learning-based scene character detector",
abstract = "In this paper, we propose a structure learning-based scene character detector which is inspired by the observation that characters have their own inherent structures compared with the background. Graphs are extracted from the thinned binary image to represent the topological line structures of scene contents. Then, a graph classifier, namely gBoost classifier, is trained with the intent to seek out the inherent structures of character and the counterparts of non-character. The experimental results show that the proposed detector achieves the remarkable classification performance with the accuracy of about 70{\%}, which demonstrates the existence and separability of the inherent structures.",
author = "Yugo Terada and Rong Huang and Yaokai Feng and Seiichi Uchida",
year = "2013",
month = "12",
day = "11",
doi = "10.1109/ICDAR.2013.101",
language = "English",
pages = "472--476",
journal = "Proceedings of the International Conference on Document Analysis and Recognition, ICDAR",
issn = "1520-5363",

}

TY - JOUR

T1 - On the possibility of structure learning-based scene character detector

AU - Terada, Yugo

AU - Huang, Rong

AU - Feng, Yaokai

AU - Uchida, Seiichi

PY - 2013/12/11

Y1 - 2013/12/11

N2 - In this paper, we propose a structure learning-based scene character detector which is inspired by the observation that characters have their own inherent structures compared with the background. Graphs are extracted from the thinned binary image to represent the topological line structures of scene contents. Then, a graph classifier, namely gBoost classifier, is trained with the intent to seek out the inherent structures of character and the counterparts of non-character. The experimental results show that the proposed detector achieves the remarkable classification performance with the accuracy of about 70%, which demonstrates the existence and separability of the inherent structures.

AB - In this paper, we propose a structure learning-based scene character detector which is inspired by the observation that characters have their own inherent structures compared with the background. Graphs are extracted from the thinned binary image to represent the topological line structures of scene contents. Then, a graph classifier, namely gBoost classifier, is trained with the intent to seek out the inherent structures of character and the counterparts of non-character. The experimental results show that the proposed detector achieves the remarkable classification performance with the accuracy of about 70%, which demonstrates the existence and separability of the inherent structures.

UR - http://www.scopus.com/inward/record.url?scp=84889565425&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84889565425&partnerID=8YFLogxK

U2 - 10.1109/ICDAR.2013.101

DO - 10.1109/ICDAR.2013.101

M3 - Conference article

AN - SCOPUS:84889565425

SP - 472

EP - 476

JO - Proceedings of the International Conference on Document Analysis and Recognition, ICDAR

JF - Proceedings of the International Conference on Document Analysis and Recognition, ICDAR

SN - 1520-5363

M1 - 6628666

ER -