Please use this identifier to cite or link to this item:
https://ptsldigital.ukm.my/jspui/handle/123456789/475552| Title: | Enhancement of moment invariant computation for off-line Arabic handwriting recognition |
| Authors: | Twana Najm Abdullah (P47937) |
| Supervisor: | Khairuddin Omar, Prof. Dr. |
| Keywords: | Enhancement Invariant computation Arabic handwriting recognition Neural networks (Computer science) |
| Issue Date: | 8-Jul-2011 |
| Description: | Moment Invariant has been frequently used as feature for shape recognition. These features are invariant to several deformations such as orientation, size and translation. The more descriptors are used, the better the characterisation of a given shape. This research proposes an enhancement of invariant based on the Flusser and Suk’s moment calculation to increase the robustness of the current descriptors. It uses the centre of an image as the replacement for the image’s “centre of gravity” that is formally used in the moment invariant calculation. The whole recognition process involves only feature extraction and classification steps. The new descriptor set was tested to recognise Arabic words based on the IFN/ENIT image database that consists of 26459 words written by 411 different writers. The back propagation Neural Network was used as the classifier. Experiment results have confirmed the basic properties of the new descriptor set and showed that the use of the proposed method of moment invariants computation had increased the recognition accuracy from 56.9% to 75.4%.,Master |
| Pages: | 56 |
| Call Number: | QA76.87.A254 2011 |
| Publisher: | UKM, Bangi |
| Appears in Collections: | Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat |
Files in This Item:
There are no files associated with this item.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.