Foreground and Background Information in an HMM-Based Method for Recognition of Isolated Characters and Numeral Strings

Authors:
Alceu de S. Britto Jr.;Robert Sabourin;Flavio Bortolozzi;Ching Y. Suen
Affiliations:
Pontifícia Universidade Católica do Paraná and Universidade Estadual de Ponta Grossa;École de Technologie Supérieure and Centre for Pattern Recognition and Machine Intelligence;Pontifícia Universidade Católica do Paraná;Centre for Pattern Recognition and Machine Intelligence
Venue:
IWFHR '04 Proceedings of the Ninth International Workshop on Frontiers in Handwriting Recognition
Year:
2004

Citing 0
Cited 5

Hybrid off-line OCR for isolated handwritten Greek characters

SPPR'07 Proceedings of the Fourth conference on IASTED International Conference: Signal Processing, Pattern Recognition, and Applications
Hybrid off-line OCR for isolated handwritten Greek characters

SPPRA '07 Proceedings of the Fourth IASTED International Conference on Signal Processing, Pattern Recognition, and Applications
Handwritten character recognition through two-stage foreground sub-sampling

Pattern Recognition
Dynamic zoning selection for handwritten character recognition

CIARP'11 Proceedings of the 16th Iberoamerican Congress conference on Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications
Off-line cursive script recognition: current advances, comparisons and remaining problems

Artificial Intelligence Review

Quantified Score

Hi-index	0.02

Visualization

Abstract

In this paper we combine complementary features based on foreground and background information in an HMM-based classifier to recognize handwritten isolated characters and numeral strings. A zoning scheme based on column and row models provides a way of dividing the character into zones without making the features size variant. This strategy allows us to avoid the character normalization, while it provides a way of having information from specific zones of the character. The experimental results on 10 digit classes, 52 character classes and 6 classes of numeral strings of different lengths have shown that the proposed features are highly discrimminant.