Turkish fingerspelling recognition with fast alignment, thick edge detection and correlation with penalization

ALTUN O., VARLI S.

IEEE 15th Signal Processing and Communications Applications Conference, Eskişehir, Türkiye, 11 - 13 Haziran 2007, ss.87-90, (Tam Metin Bildiri) identifier identifier

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Doi Numarası: 10.1109/siu.2007.4298841
  • Basıldığı Şehir: Eskişehir
  • Basıldığı Ülke: Türkiye
  • Sayfa Sayıları: ss.87-90
  • Yıldız Teknik Üniversitesi Adresli: Evet

Özet

Fingerspelling is used for words that have no sign, or for words that the sign is unknown. In this work, the signs of Turkish Alphabet are recognized from video images. The methods used are fast alignment, thick edge extraction and correlation with penalization methods. A 99 percent success ratio is achieved.