Using a hidden Markov model to transcribe handwritten bushman texts

Authors:
Kyle Williams;Hussein Suleman
Affiliations:
University of Cape Town, Cape Town, South Africa;University of Cape Town, Cape Town, South Africa
Venue:
Proceedings of the 11th annual international ACM/IEEE joint conference on Digital libraries
Year:
2011

Citing 5
Cited 1

On the Influence of Vocabulary Size and Language Models in Unconstrained Handwritten Text Recognition

ICDAR '01 Proceedings of the Sixth International Conference on Document Analysis and Recognition
Offline Recognition of Unconstrained Handwritten Texts Using HMMs and Statistical Language Models

IEEE Transactions on Pattern Analysis and Machine Intelligence
Discrete Cosine Transfom

IEEE Transactions on Computers
Off-line recognition of realistic Chinese handwriting using segmentation-free strategy

Pattern Recognition
Digital libraries without databases: the Bleek and Lloyd collection

ECDL'07 Proceedings of the 11th European conference on Research and Advanced Technology for Digital Libraries

A comparison of machine learning techniques for handwritten |Xam word recognition

Proceedings of the South African Institute for Computer Scientists and Information Technologists Conference

Quantified Score

Hi-index	0.01

Visualization

Abstract

The Bushman texts in the Bleek and Lloyd Collection contain complex diacritics that make automatic transcription difficult. Transcriptions of these texts would allow for enhanced digital library services to be created for interacting with the collection. In this study, an investigation into automatic transcription of the Bushman texts was performed using the popular method of using a Hidden Markov Model for text line recognition. The results show that while this technique may be well suited to well-constrained and understood scripts, its application to more complex scripts introduces a number of difficulties that need to be overcome.