Dil Nawaz Hakro

Printed Text Image Database for Sindhi OCR

Document Image Understanding (DIU) and Electronic Document Management are active fields of research involving image understanding, interpretation, efficient handling, and routing of documents as well as their retrieval. Research on most of the noncursive scripts (Latin) has matured, whereas research on the cursive (connected) scripts is still moving toward perfection. Many researchers are currently working on the cursive scripts (Arabic and other scripts adopting it) around the world so that the difficulties and challenges in document understanding and handling of these scripts can be overcome. Sindhi script has the largest extension of the original Arabic alphabet among languages adopting the Arabic script; it contains 52 characters, compared to 28 characters in the original Arabic alphabet, in order to accommodate more sounds for the language. There are 24 differentiating characters with some possessing four dots. For Sindhi OCR research and development, a database is needed for training and testing of Sindhi text images. We have developed a large database containing over 4 billion words and 15 billion characters in 150 various fonts in four font weights and four styles. The database contents were collected from various sources including websites, books, and theses. A custom-built application was also developed to create a text image from a text document that supports various fonts and sizes. The database considers words, characters, characters with spaces, and lines. The database is freely available as a partial or full database by sending an email to one of the authors.

Phonetic-based Sindhi spellchecker system using a hybrid model

This article presents a novel architecture using a hybrid model for developing a Sindhi spellchecker system which has yet not been developed prior to this work. The compound textual forms and glyphs of Sindhi language presents a substantial challenge for developing a Sindhi spellchecker system and generating a similar suggestion list for misspelled words. In order to implement such a system, phonetic-based Sindhi language rules and patterns must be taken into account for increasing the accuracy and efficiency. In this research work, a simple and efficient combinational hybrid system is proposed, using three different algorithms, the Edit Distance algorithm to find the measure of similarity between two Sindhi strings. The phonetic-based SoundEx and ShapeEx algorithms are developed for pattern or glyph matching, generating accurate and an efficient suggestion list for incorrect or misspelled Sindhi words. The proposed system is established with a blend between Phonetic-based SoundEx algorithm and ShapeEx algorithm for pattern or glyph matching, generating accurate and efficient suggestion list for incorrect or misspelled Sindhi words. In this article, a table of phonetically similar-sounding Sindhi characters is presented which are grouped together along with another table containing similar glyph or shape-based character groups. The system has been successfully integrated into a pre-developed Sindhi word processer application. The Sindhi word segmentation methodology and algorithms required for the spellchecker has already been published and so are not discussed in detail in this article.