Arabic OCR evaluation tool
Authors:
- 25 August 2016
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Abstract
Performance evaluation of Optical Character Recognition (OCR) systems is an essential task for OCR systems development. However, studies in Arabic OCR suffer from the lack of proper performance evaluation metrics and the availability of evaluation tools. Although the literature provides typical performance metrics, such as character accuracy and word accuracy for OCR performance evaluation, these metrics are not sufficient for evaluating Arabic OCR. This paper presents an open source automated software tool with various metrics for the evaluation of Arabic OCR performance. The developed tool is available for OCR researchers, thus it can be applied for ranking different OCR algorithms.Keywords
This publication has 17 references indexed in Scilit:
- Open-vocabulary recognition of machine-printed Arabic text using hidden Markov modelsPattern Recognition, 2016
- Primitive printed Arabic Optical Character Recognition using statistical features2015
- Robust metrics for evaluating arabic OCR systems2014
- An open-source OCR evaluation tool2014
- A study on font-family and font-size recognition applied to Arabic word images at ultra-low resolutionPattern Recognition Letters, 2013
- Software tools and test data for research and testing of page-reading OCR systems2005
- A corpus for comparative evaluation of OCR software and postcorrection techniques2005
- Performance metrics for document understanding systems2002
- Correcting English text using PPM models2002
- Performance evaluation of two Arabic OCR products1999