Segmentation-free word recognition with application to Arabic

Abstract

This paper describes the desagn and amplementataon of a system that recognazes machzne-pranted Arabic words without praor segmentataon. The technaque as based on describang symbols an terms of shape praniatives. A t recognataon tame, the pramataves are detected on a word amage usang mathematacal morphology operataons. The system then matches the detected priniataves wath symbol models. Thas leads to a spataal arrangement of matched symbol models. The system conducts a search an the space of spataal arrangements of models and outputs the arrangement wath the liaghest posterior probabilaty as the recognataon of the word.The advantage of usang thas whole word approach versus a segmentataon approach as that the result of recognation is optimazed wath regard to the whole word. Results of preliminary experiments usang a letacon of 42,OOO words show a recognataon rate of 99.4% for noise-free text and 73% for scanned tez.

A Recognition-Based Approach to Segmenting Arabic Handwritten Text

“…Adjacent letters that form shapes similar to those of classes 15,16,17,18,19,20,21 and 22 may wrongly be combined although they are correctly classified before grouping. This scenario is very clear in Figure 9, in the word ‫,"ﺳﯿﺪي"‬ where the adjacent letters ‫"ﯿ"‬ and part of the letter ‫"ﺳ"‬ were recognized and classified as letter ‫,"ﺳ"‬ and the remaining part of ‫"ﺳ"‬ was classified as letter ‫.…”

“…In 2000, Amin introduced another holistic approach where global features such as loops and peaks were extracted from the input word [17], and passed to the C4.5 machine learning system to generate a decision tree for classifying the word. The success rate of the system was 92% using 1000 Arabic words with different fonts.…”

A Front-End OCR for Omni-Font Persian/Arabic Cursive Printed Documents

“…Thereafter, the problem of understanding of the script reduces into recognition of the words. There are two main approaches to automatic understanding of cursive scripts: holistic and segmentation-based [4]. In the first approach, each word is treated as a whole and the recognition system does not consider it as a combination of separable characters.…”

A segmentation-free approach to Arabic and Urdu OCR

“…A lot of studies have been done in order to investigate the challenges associated with recognition for Arabic script languages, and to suggest approaches that take into account these challenges. Most of the research done in Arabic script OCR is mainly for the Arabic language, such as Badr et al, 1 Khorsheed, 2 and Cheung et al 3 However, research for other Arabic script languages such as Persian and Urdu appeared much later, and is even more limited. Research for Urdu language recognition has recently started to grow, e.g., Pal et al 4 and Hussain et al 5 As for the available OCR products, a small number of commercial systems provide recognition for some Arabic script languages.…”