Paper Type |
Opinion |
Title |
Digits Recognition with Moment Invariants |
Author |
Siti Mariyam Hj. Shamsuddin[a], Md. Nasir Sulaiman[a], and Maslina Darus[b] |
Email |
- |
Abstract: This paper introduces an experimental evaluation of the utilizing various moments order as pattern features in recognition of handprinted and handwritten digits. The moments that have been used are Geometric moments and Aspect invariant moments. These moments have been used as feature extraction for digits with various orientations and scaling. We use standard backpropagation and an improved backpropagation for the classification of handprinted and handwritten digits for comparison. We found that the performance of handprinted and handwritten digits classifications are dependent on the moment order. Experiments on those digits using different moments order are carried out in classification phase, and the results are promising for moments of higher order.
|
|
Start & End Page |
46 - 56 |
Received Date |
1999-10-27 |
Revised Date |
|
Accepted Date |
2000-01-28 |
Full Text |
Download |
Keyword |
moment invariants, digit recongnition, backpropagation, feature extraction |
Volume |
Vol.27 No.1 (JUNE 2000) |
DOI |
|
Citation |
Shamsuddin S.M.H., Sulaiman M.N. and Darus M., Digits Recognition with Moment Invariants, Chiang Mai J. Sci., 2000; 27(1): 46-56. |
SDGs |
|
View:618 Download:174 |