Recent approaches in online handwriting recognition for Persian and Arabic right-to-left languages

Abstract

Online handwriting recognition redefined along the rapid change of technology which began to replace keyboards with digital pen and tablets. However compared to English letter-based languages, less attention is paid to right-to-left languages especially Persian and Arabic. The complex nature of cursive writing in these languages acquires elaborate methods to simultaneously take speed, accuracy and complexity into account. This paper overviews the up-to-date efforts put into online Persian and Arabic handwriting recognition. Through this paper word/stroke-based and segment-based methods are discussed. Additionally template based recognition algorithms and artificial intelligent approaches are explained with respect to dictionary check requirements and benchmark databases. Real-time online handwriting recognition of such languages is another recent field of interest that is researched for potentially practical methods.