Efficient off-line verification and identification of signatures by multiclass support vector machines
11th International Conference on Computer Analysis of Images and Patterns, CAIP 2005, Versailles, Fransa, 5 - 08 Eylül 2005, ss.799-805, (Tam Metin Bildiri)
- Yayın Türü: Bildiri / Tam Metin Bildiri
- Cilt numarası:
- Doi Numarası: 10.1007/11556121_98
- Basıldığı Şehir: Versailles
- Basıldığı Ülke: Fransa
- Sayfa Sayıları: ss.799-805
- Yıldız Teknik Üniversitesi Adresli: Evet
Özet
In this paper we present a novel and efficient approach for off-line signature verification and identification using Support Vector Machine. The global, directional and grid features of the signatures were used. In verification, one-against-all strategy is used. The true acceptance rate is 98% and true rejection rate is 81%. As the identification of signatures represent a multi-class problem, Support Vector Machine's one-against-all and one-against-one strategies were applied and their performance were compared. Our experiments indicate that one-against-one with 97% true recognition rate performs better than one-against-all by 3%. © Springer-Verlag Berlin Heidelberg 2005.