| LEADER |
02025nam a22002057a 4500 |
| 001 |
0008902 |
| 041 |
|
|
|a eng
|
| 044 |
|
|
|b المغرب
|
| 100 |
|
|
|9 29965
|a Fakir, Mohamed
|e Author
|
| 245 |
|
|
|a Automatic recognition of arabic printed scripts
|
| 260 |
|
|
|b معهد الدراسات والأبحاث للتعريب
|c 1997
|g أكتوبر
|
| 300 |
|
|
|a 121 - 146
|
| 336 |
|
|
|a بحوث المؤتمرات
|b Article
|
| 520 |
|
|
|b The subject of character recognition has been receiving considerable attention in recent years due to the increasing dependence on computer data processing. Several methods for recognizing Latin, Chinese and Arabic characters have been proposed. In this paper, a recognition system for Arabic printed scripts is presented. The process consists of three phases. In the first phase, preprocessing is done. In the second phase, features are extracted using Hough transform (HT),. Fourier descriptors (FD) and moment invariants. In the last one, characters are classified in two steps: Firstly, the character main body is classified using Dynamic Programming (DP). Secondly, simple topological features extracted from the geometry of the stress marks are used by the topological classifier to completely recognize the character. The result obtained by HT application is compared to the ones obtained by using Fourier descriptors and moment invariants. By the application of moment invariants, characters are recognized in one step. It has been found that moment invariants give high recognition rate than HT and FD.
|
| 653 |
|
|
|a اللغة العربية
|a الحروف العربية
|a تكنولوجيا المعلومات
|a المطبوعات
|a معالجة البيانات
|
| 773 |
|
|
|c 005
|d الرباط
|i معهد الدراسات و الأبحاث للتعريب بالرباط
|l 000
|m مج2
|o 6876
|s أعمال الندوة الدولية: تمكين اللغات، التكنولوجيا والاتصال
|v 002
|
| 856 |
|
|
|u 6876-002-000-005.pdf
|
| 930 |
|
|
|d y
|p y
|
| 995 |
|
|
|a AraBase
|
| 999 |
|
|
|c 594242
|d 594242
|
