Optical Character Recognition for Arabic Ruq

In the ongoing efforts in Operation Iraqi Freedom (OIF) and Operation Enduring Freedom (OEF), as well as the continuing Global War on Terror (GWOT), U.S. forces gain valuable information from captured documents. While these documents are typically written in Arabic, there are not enough U.S. personnel trained in reading Arabic to expediently read and understand the captured information. Since the information has the most value while it is still fresh, the U.S. military has started to use Arabic based optical character recognition (OCR) to quickly convert the Arabic text into English. While these commercial OCR packages work well with machine-produced documents, their performance is quickly and significantly degraded on more typical hand-written Arabic documents. The problems are compounded when the original document contains even small levels of noise or other sources of image degradation. In order to more quickly and effectively process collected information there is a need for an automated system to process this handwritten text. We propose to leverage our experience with developing previous image processing and handwriting recognition systems to develop a system for performing optical character recognition (OCR) on the most commonly used Arabic script, Ruq’ah