Arabic Handwriting Recognition Solution

Description

A new on-line handwriting recognition system for Arabic personal names based on Hidden Markov Model (HMM). The system is trained with the ADAB-database using two different methods: manually segmented characters and non-segmented words. The system is to be used for touch screens that recognize Arabic writing. A set of around 70,000 personal names were collected from the databases of different organizations. This set was reduced to remove the repeated names and resulted into 2,800 different Arabic personal names that represent the vocabulary of this system. The system is trained with the ADAB-database using two different methods: manually segmented characters and non-segmented words. The recognition system that deals with a large vocabulary of Arabic personal names uses a new lexicon reduction method that depends on the delayed strokes formation and the number of strokes. The new delayed strokes detection method is used to reduce the temporal variation of the on-line sequence. A dataset of on-line Arabic handwritten names has been collected to validate the system and a highly encouraging recognition rate was achieved compared to the results of commercially available recognition systems on the same dataset. The system is trained using the on-line database ADAB of Tunisian town names that is divided into 3 sets containing 23,251 words (122,559 characters).

Aspects

The technology acts as a fast system for Arabic handwriting recognition that is highly accurate and efficient. The system has minimal processing footprint.

Patenting Status

Pending

Status

Patented

Target Business

License Agreement

Submit Your Intellectual Property Send Your Feedback