Title :
A new point of view to common vector approach
Author :
Gulmezoglu, M. Bilginer ; Çevikalp, Hakan ; Barkana, Atalay
Author_Institution :
Osmangazi Univ., Eskisehir, Turkey
Abstract :
The common vector approach is handled from a new point of view. Thus, all classes can be represented by a unique subspace and high recognition rates can be obtained by using feature vectors with very low dimensions. The recognition rates obtained from an experimental study using the TI-digit database are also given.
Keywords :
pattern recognition; vectors; common vector approach; feature vectors; pattern recognition; recognition rates; Spatial databases; Testing;
Conference_Titel :
Signal Processing and Communications Applications Conference, 2004. Proceedings of the IEEE 12th
Print_ISBN :
0-7803-8318-4
DOI :
10.1109/SIU.2004.1338635
