Script based trilingual handwritten word level multiple skew estimation

Ravikumar, M. and Guru, D. S. and Manjunath, S. and Manjunath Aradhya, V. N. (2016) Script based trilingual handwritten word level multiple skew estimation. In: Information Systems Design and Intelligent Applications. Springer India, New Delhi, pp. 541-549. ISBN 978-81-322-2752-6

Full text not available from this repository. (Request a copy)
Official URL: https://doi.org/10.1007/978-81-322-2752-6_53

Abstract

Skew estimation and correction plays an important role in document analysis. In the present work, we proposed a model to estimate multiple skews present in trilingual such as Devanagari, English, and Kannada handwritten documents at word level with a priori knowledge about the corresponding scripts. The idea of using different skew estimation techniques for different scripts such as Hough transform (HT) for Devanagari words, Gaussian Mixture Models (GMM) and convex hull for Kannada and English words is proposed. The effectiveness of these approaches has been reported by testing on a dataset consisting of 1000 words in each script. Experimental results show that the proposed approaches are effective in estimating and correcting the handwritten skew words.

Item Type: Book Section
Additional Information: Unmapped bibliographic data: ED - Satapathy, Suresh Chandra [Field not mapped to EPrints] ED - Mandal, Jyotsna Kumar [Field not mapped to EPrints] ED - Udgata, Siba K. [Field not mapped to EPrints] ED - Bhateja, Vikrant [Field not mapped to EPrints] BT - Information Systems Design and Intelligent Applications: Proceedings of Third International Conference INDIA 2016, Volume 2 [Field not mapped to EPrints]
Subjects: Information, Computer and Applied Sciences > Computer Science
Divisions: PG Campuses > Manasagangotri, Mysore > Computer Science
Depositing User: Praveen Kumari B.L
Date Deposited: 26 Oct 2017 05:36
Last Modified: 26 Oct 2017 05:36
URI: http://eprints.uni-mysore.ac.in/id/eprint/20102

Actions (login required)

View Item View Item