A Complete Off-line Sindhi Handwritten Text Recognition: A Survey
Shafique Ahmad Awan, Dil Nawaz Hakro, Intzar Lashari, Akhtar Hussain Jalbani, Maryam Hameed · Zenodo (CERN European Organization for Nuclear Research) · 2017
Artificial Intelligence is finding ways to make machines more intelligent and work like human being. Image processing, Natural language processing and Optical Character Recognition (OCR) are the active fields of computer vision, where the computers are made more versatile to understand, read and write natural human languages spoken around the word. Optical Characters Recognition (OCR) and Intelligent Characters Recognition (ICR) differ in recognizing printed and handwritten characters respectively. Intelligent Characters Recognition (ICR) is an active field in which handwritten characters are converted into editable text from the image, and remain the point of interest for researchers around the world. Many of the languages of the world possess their Intelligent Characters Recognition (ICR) or their ICR systems are in process. Latin scripts possess their ICR and are near to perfect whereas Arabic script and its adopting languages need more attention for the development of ICR systems. Sindhi language is a language having rich background and culture of more than 5000 years still lacks the ICR system. As there is no any handwritten recognition system for Sindhi Language, so there is no handwritten database is available for testing and training. Enhanced segmentation and feature extraction algorithms are needed which can fully suit with Sindhi script. An integrated handwritten system will be the output of this system in which handwritten text is recognized and editable text will be available for the further processing.