In this paper, we review classic handwriting recognition processes and controversial issues by details. In addition, since there are presented the state of the art approaches based on neural networks solutions, we review RNN approaches. Furthermore, as the database is an important factor in RNN based solutions, we survey the most important datasets for the left to right and right to left writing languages. In the following, we have presented a systematic survey of some important datasets applications and most important presented handwriting recognition approaches.
Keywords:
Handwriting Recognition, Deep Learning, OCR, Synthesized Data.
