A first-of-its-kind system for real-time Arabic handwriting recognition was developed at Ben-Gurion University

The innovative development has a far-reaching impact on about a billion people

Researchers at Ben-Gurion University of the Negev succeeded in developing a first-of-its-kind system for identifying Arabic handwriting in real time with a very high level of accuracy - about 97%. The system allows the user to write with a pen on an electronic board, which feeds the handwriting to a computer, which analyzes and translates it into printed letters in milliseconds. The development was recently registered as a patent.

The researchers, Dr. Gihad Al-Sanae and his master's student Fadi Bayadasa from the University's Computer Science Department, developed the system based on research based on advanced learning methods, using computational geometry to analyze the curves, loops and points that build a probabilistic model of identification . The system recognizes average handwriting and is built to learn new handwritings during writing.

Mouse and keyboard are the most common devices for transferring textual information to a computer. The development of handheld computers and personal tablets, which allow writing by hand on a small board, requires new methods for recognizing handwriting and translating it into printed text on a computer.

Dr. El Sane: "When the use of the computer becomes an important element in providing service to the citizen, there is a huge advantage to using handwriting instead of typing, especially in developing countries. Today there are systems for recognizing Latin handwriting in good percentages, but there is no practical system for recognizing Arabic handwriting, which may be used by about six of the inhabitants of the planet. Identifying Arabic handwriting is a difficult problem, since this writing is naturally connected and because most Arabic letters contain dots that are placed above or below the letter. It will be emphasized that we commit to the same form of natural writing without change. For example, in some handheld computers, even the Latin letters have a small change in the letters, so that they can be recognized quickly and without errors."

As for the economic potential of the development, Dr. Al Sana said: "It is a target population of about one billion people who speak the Arabic language, the Urdu language (Pakistan) and Kurdish, who may be potential users of the development. As far as we are concerned, the sky is the limit." Ben-Gurion University's implementation company, B.G. Negev Technologies is currently working to locate investments in the project.