We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Text Restoration Model Using OCR and Spelling Correction Algorithms.
- Authors
Liza Ramos, Anna; Mina, Toblerone; Mangilaya, Jeremy; Lapasa, Donnalyn T.; Manaig, Elizabeth; Luiz Bagacina, John Ace
- Abstract
Printed documents are valuable assets of information and must be preserved for future reference. This study aims to restore text in different types of degradation, apply a post-processing techniques to increase accuracy and to address the limitation of OCR in misspelled words. The model consists of 500 words written in Times New Roman, Calibri, Helvetica and Arial. Results showed wet documents achieved the highest accuracy rate among other types of degradation with a score of 99.2% while the ones written over with pens achieved 91.2% for most of the conjoined words caused by the degradation were not recognized by the model. Arial fonts had the best results while Times New Roman acquired the lowest performance rate. In terms of spelling corrections, the model showed impressive results, correcting more than half of the misspelled words.
- Subjects
OPTICAL character recognition; SPELL checkers (Computer programs); SPELLING errors; ALGORITHMS; TEXT recognition; LEGIBILITY (Printing)
- Publication
International Journal of Simulation -- Systems, Science & Technology, 2019, Vol 20, p35.1
- ISSN
1473-8031
- Publication type
Article
- DOI
10.5013/IJSSST.a.20.S2.35