VML-MOC: Segmenting a multiply oriented and curved handwritten text line dataset

Berat Kurar Barakat; Rafi Cohen; Jihad El-Sana; Irina Rabaev

doi:10.1109/ICDARW.2019.50109

VML-MOC: Segmenting a multiply oriented and curved handwritten text line dataset

Berat Kurar Barakat, Rafi Cohen, Jihad El-Sana, Irina Rabaev

Department of Computer Science

Research output: Contribution to journal › Article › peer-review

Abstract

This paper publishes a natural and very complicated dataset of handwritten documents with multiply oriented and curved text lines, namely VML-MOC dataset. These text lines were written as remarks on the page margins by different writers over the years. They appear at different locations within the orientations that range between 0° and 180° or as curvilinear forms. We evaluate a multi-oriented Gaussian based method to segment these handwritten text lines that are skewed or curved in any orientation. It achieves a mean pixel Intersection over Union score of 80.96% on the test documents. The results are compared with the results of a single-oriented Gaussian based text line segmentation method.

Original language	English
Pages (from-to)	13-18
Journal	Proceedings of the International Conference on Document Analysis and Recognition, ICDAR
DOIs	https://doi.org/10.1109/ICDARW.2019.50109
State	Published - Nov 2019

Access to Document

10.1109/ICDARW.2019.50109

Cite this

@article{d5cbf4159a00425c973cfeede915f27b,

title = "VML-MOC: Segmenting a multiply oriented and curved handwritten text line dataset",

abstract = "This paper publishes a natural and very complicated dataset of handwritten documents with multiply oriented and curved text lines, namely VML-MOC dataset. These text lines were written as remarks on the page margins by different writers over the years. They appear at different locations within the orientations that range between 0° and 180° or as curvilinear forms. We evaluate a multi-oriented Gaussian based method to segment these handwritten text lines that are skewed or curved in any orientation. It achieves a mean pixel Intersection over Union score of 80.96% on the test documents. The results are compared with the results of a single-oriented Gaussian based text line segmentation method.",

author = "Barakat, {Berat Kurar} and Rafi Cohen and Jihad El-Sana and Irina Rabaev",

year = "2019",

month = nov,

doi = "10.1109/ICDARW.2019.50109",

language = "English",

pages = "13--18",

journal = "Proceedings of the International Conference on Document Analysis and Recognition, ICDAR",

issn = "1520-5363",

publisher = "Institute of Electrical and Electronics Engineers",

}

TY - JOUR

T1 - VML-MOC

T2 - Segmenting a multiply oriented and curved handwritten text line dataset

AU - Barakat, Berat Kurar

AU - Cohen, Rafi

AU - El-Sana, Jihad

AU - Rabaev, Irina

PY - 2019/11

Y1 - 2019/11

N2 - This paper publishes a natural and very complicated dataset of handwritten documents with multiply oriented and curved text lines, namely VML-MOC dataset. These text lines were written as remarks on the page margins by different writers over the years. They appear at different locations within the orientations that range between 0° and 180° or as curvilinear forms. We evaluate a multi-oriented Gaussian based method to segment these handwritten text lines that are skewed or curved in any orientation. It achieves a mean pixel Intersection over Union score of 80.96% on the test documents. The results are compared with the results of a single-oriented Gaussian based text line segmentation method.

AB - This paper publishes a natural and very complicated dataset of handwritten documents with multiply oriented and curved text lines, namely VML-MOC dataset. These text lines were written as remarks on the page margins by different writers over the years. They appear at different locations within the orientations that range between 0° and 180° or as curvilinear forms. We evaluate a multi-oriented Gaussian based method to segment these handwritten text lines that are skewed or curved in any orientation. It achieves a mean pixel Intersection over Union score of 80.96% on the test documents. The results are compared with the results of a single-oriented Gaussian based text line segmentation method.

U2 - 10.1109/ICDARW.2019.50109

DO - 10.1109/ICDARW.2019.50109

M3 - Article

SN - 1520-5363

SP - 13

EP - 18

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

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

ER -

VML-MOC: Segmenting a multiply oriented and curved handwritten text line dataset

Abstract

Access to Document

Fingerprint

Cite this