Ottoman Handwriting

SUPPORTING OF THE PROJECT 'IDENTIFICATION OF OTTOMAN HANDWRITTEN DOCUMENTS BY USING SMART SYSTEMS'

The project entitled as 'Translation of Printed/Handwritten Ottoman texts into Latin letter Turkish by using artificial intelligence based image processing techniques and smart database' led by Prof. Dr. Osman Nuri Uçan and Prof. Dr. Ahmet Kala in order to accelerate benefiting from 16th century and later sources of Turkish Culture was supported by the Elginkan Foundation and was found successful and concluded in 20.05.2009.

In our project that continued for 12 months, it was aimed to translate Ottoman texts into Latin letter Turkish in order to accelerate benefiting from sources of Turkish Culture.

The fact that Ottoman has almost 2.000.000 documents, is used in daily life and lack of experts who will evaluate according to objective criteria, increases the importance of the software.

Artificial intelligence based image processing techniques such as Artificial Neural Networks (ANN) and wavelet method were used in solving complex problems.

Use of smart database was generated in a manner that will minimize the errors that will occur in identification of letters and syllables, in particular.  An all-around original dictionary, in which letters from A to Z, words, verbs and conjugations, and vowels of each word are specified, were generated.

The software was completed as a whole and applied on an exemplary page of Wealth of Sciences magazine.

Intermediate results of our project, which was executed at international quality standards, were published in a respectable magazine within SCI. Also a presentation was made in an international conference in Malta and the conference organized in our country by the Chamber of Electrical and Electronics Engineering.