 |
A Student from Computer Science Department get PhD about identifying handwritten text for security application |
A Student from Computer Science Department get PhD about identifying handwritten text for security application
The committee of discussion consisted of:-
1- Prof.Dr. Azhar Hadi Abdullah / Diyala University / Computer Science Department.
2- Assist.Prof.Dr. Haitham Omar Latif /Al-Nahreen University / Computer Science Department
3- Assist.Prof.Dr Shayma Shakir Hamid and Assist.Prof.Dr Aliaa Abdul Karim and Assist.Prof.Dr. Abdul Ameer Abdullah and Assist.Prof.Dr Mtheel Imad Eddin Computer Science Department / University of technology
During the search it has been suggested an accurate and reliable system to identify handwritten Arabic text on the basis of parts the inputs from handwritten documents into sub handwritten words. The system contains two basis parts (modules) that are used to identify the handwritten text and identify the writer. The first part (module1) has six stages that working together to know Arabic text handwriting and convert it to adjustable text. These stages are: acquiring images, retail, processing, construction of base features, classification and stage after treatment. While the second part (module2) selects the desired document book through several stages
Which are: Aktas images, retail, processing, features extraction, classification and post-processing
The algorithm system proposed an effective and accurate segmentation that part handwritten text entrance into a number of sub-images that written and each sub-image containing Arabic subset word. In addition it has been suggested a threshold algorithm for converting sub-images to binary image by using the aggregation function (fuzzy c-mean). Moreover, sub- bilateral images pass through algorithm proposed to remove Alaudhae in order to remove unwanted pixels.
In order to test the system's performance, it had been used three statements of Arabic language handwritten rules, which is a database
AHDB, database IESK-arDB data and proposed data base of Arabic language handwritten.
