Survey of Multilingual Script Identification Techniques on Wild Images

Authors

  • Kiran Perveen Department Computer Science, COMSATS Institute of Information Technology (CIIT), Islamabad. https://orcid.org/0000-0002-1158-369X
  • Rukhsana Perveen Department of Electrical Engineering, National University of Computer and Emerging Sciences (FAST), Islamabad. https://orcid.org/0000-0001-5244-1463
  • Awais Yasin Department of Electrical Engineering, National University of Technology (NUTECH), Islamabad-Pakistan.

DOI:

https://doi.org/10.5281/zenodo.6547188

Keywords:

Latest script, wild images, style, script techniques, challenges images

Abstract

Multilingual Script Identification on natural images has recently increase research attention and this is very challenging task. This paper presents a review of latest techniques for the multilingual scripts. The system can choose the appropriate Optical Character Recognition (OCR) engine to recognize a script based here on script identity of a retrieved line of text or word. A number of approaches for identifying different characters, including such Japanese, Chinese, Arabic, Korean and Indian, have been developed. scripts are used in written on natural scenes captured by a voyager from cameras or text recognitions system. Here we also present the difficulties that come with script identification, methods used for features extraction and also the classifiers used for identification. We provided a comprehensive description and evaluation of previous and state-of-the-art script identification approaches. It should be emphasized that researchers in the area of multilingual script recognition is still in its early stages, and additional analysis is needed.

Downloads

Download data is not yet available.

Downloads

Published

2022-04-06

How to Cite

Perveen, K., Perveen, R. ., & Yasin, D. A. . (2022). Survey of Multilingual Script Identification Techniques on Wild Images. LC International Journal of STEM (ISSN: 2708-7123), 3(1), 1-14. https://doi.org/10.5281/zenodo.6547188