An HMM Based Two-Pass Approach for Off-Line Cursive Handwriting Recognition

Authors:
Wenwei Wang;Anja Brakensiek;Gerhard Rigoll
Affiliations:
-;-;-
Venue:
ICMI '00 Proceedings of the Third International Conference on Advances in Multimodal Interfaces
Year:
2000

Citing 3
Cited 0

HMM Word Recognition Engine

ICDAR '97 Proceedings of the 4th International Conference on Document Analysis and Recognition
Performance Evaluation of a New Hybrid Modeling Technique for Handwriting Recognition Using Identical On-Line and Off-Line Data

ICDAR '99 Proceedings of the Fifth International Conference on Document Analysis and Recognition
Modeling and recognition of cursive words with hidden Markov models

Pattern Recognition

Quantified Score

Hi-index	0.00

Visualization

Abstract

The cursive handwriting recognition is a challenging task because the recognition system has to handle not only large shape variation of human handwriting, but also character segmentation. Usually the recognition performance depends crucially upon the segmentation process. Hidden Markov Models (HMMs) have the ability to model similarity and variation among samples of a class. In this paper we present an extended sliding window feature extraction method and an HMM based two-pass modeling approach. Whereas our feature extraction method makes the resulting system more robust with word baseline detection, the two-pass recognition approach exploits the segmentation ability of the Viterbi algorithm and creates another HMM set and carries out a second pass recognition. The total performance is enhanced by combination of the two pass results. Experiments of recognizing cursive handwritten words with 30000 words lexicon have been carried out and show that our novel approach can achieve better recognition performance and reduce the relative error rate significantly.