scholarly journals Arabic Cursive Text Recognition from Natural Scene Images

2019 ◽  
Vol 9 (2) ◽  
pp. 236 ◽  
Author(s):  
Saad Ahmed ◽  
Saeeda Naz ◽  
Muhammad Razzak ◽  
Rubiyah Yusof
Keyword(s):  
Recognition System ◽  
Document Image ◽  
Text Recognition ◽  
Chinese Script ◽  
Challenging Problem ◽  
Future Directions ◽  
Scene Text ◽  
Comprehensive Survey ◽  
Recognition Systems ◽  
Scene Text Recognition

This paper presents a comprehensive survey on Arabic cursive scene text recognition. The recent years’ publications in this field have witnessed the interest shift of document image analysis researchers from recognition of optical characters to recognition of characters appearing in natural images. Scene text recognition is a challenging problem due to the text having variations in font styles, size, alignment, orientation, reflection, illumination change, blurriness and complex background. Among cursive scripts, Arabic scene text recognition is contemplated as a more challenging problem due to joined writing, same character variations, a large number of ligatures, the number of baselines, etc. Surveys on the Latin and Chinese script-based scene text recognition system can be found, but the Arabic like scene text recognition problem is yet to be addressed in detail. In this manuscript, a description is provided to highlight some of the latest techniques presented for text classification. The presented techniques following a deep learning architecture are equally suitable for the development of Arabic cursive scene text recognition systems. The issues pertaining to text localization and feature extraction are also presented. Moreover, this article emphasizes the importance of having benchmark cursive scene text dataset. Based on the discussion, future directions are outlined, some of which may provide insight about cursive scene text to researchers.

scholarly journals OCR with the Deep CNN Model for Ligature Script-Based Languages like Manchu

2021 ◽  
Vol 2021 ◽  
pp. 1-9
Author(s):  
Diandian Zhang ◽  
Yan Liu ◽  
Zhuowei Wang ◽  
Depei Wang
Keyword(s):  
Recognition Accuracy ◽  
Sliding Window ◽  
Recognition System ◽  
Text Retrieval ◽  
Text Recognition ◽  
Manual Segmentation ◽  
Low Resource ◽  
Sample Data ◽  
Deep Cnn

Manchu is a low-resource language that is rarely involved in text recognition technology. Because of the combination of typefaces, ordinary text recognition practice requires segmentation before recognition, which affects the recognition accuracy. In this paper, we propose a Manchu text recognition system divided into two parts: text recognition and text retrieval. First, a deep CNN model is used for text recognition, using a sliding window instead of manual segmentation. Second, text retrieval finds similarities within the image and locates the position of the recognized text in the database; this process is described in detail. We conducted comparative experiments on the FAST-NU dataset using different quantities of sample data, as well as comparisons with the latest model. The experiments revealed that the optimal results of the proposed deep CNN model reached 98.84%.

The A2iA Multi-lingual Text Recognition System at the Second Maurdor Evaluation

2014 ◽  
Author(s):  
Bastien Moysset ◽  
Theodore Bluche ◽  
Maxime Knibbe ◽  
Mohamed Faouzi Benzeghiba ◽  
Ronaldo Messina ◽  
...  
Keyword(s):  
Recognition System ◽  
Text Recognition

scholarly journals Recognizing Cursive Typewritten Text Using Segmentation-Free System

2015 ◽  
Vol 2015 ◽  
pp. 1-7 ◽  
Author(s):  
Mohammad S. Khorsheed
Keyword(s):  
Feature Extraction ◽  
Feature Vector ◽  
Feature Space ◽  
Recognition System ◽  
Image Feature ◽  
Text Recognition ◽  
Statistical Features ◽  
Free System ◽  
Novel Method ◽  
Image Pixels

Feature extraction plays an important role in text recognition as it aims to capture essential characteristics of the text image. Feature extraction algorithms widely range between robust and hard to extract features and noise sensitive and easy to extract features. Among those feature types are statistical features which are derived from the statistical distribution of the image pixels. This paper presents a novel method for feature extraction where simple statistical features are extracted from a one-pixel wide window that slides across the text line. The feature set is clustered in the feature space using vector quantization. The feature vector sequence is then injected to a classification engine for training and recognition purposes. The recognition system is applied to a data corpus which includes cursive Arabic text of more than 600 A4-size sheets typewritten in multiple computer-generated fonts. The system performance is compared to a previously published system from the literature with a similar engine but a different feature set.

Sign in / Sign up

Export Citation Format

Share Document