AR2LT

SUMMARY

With more than 313 Million speaker around the world and being the fifth most spoken

language in the world, Arabic language indeed is popular, but unfortunately Arabic is still

falling behind in the world of technology, therefore being native speakers of the language , we

have decided to base our project mainly on developing an application which supports Arabic

OCR , the application main idea is to capture Arabic words from images as input and

transliterate to Latin characters as output, this shall be done using image processing techniques

which will process the input image and segment each letter using a segmentation algorithm

we designed after that we obtain individual sperate letters in the final step a neural network

model is used to predict the letter and give the transliterated Latin character .