Batman Universitesi Yasam Bilimleri Dergisi
www.yasambilimleridergisi.com
Cilt 1, Sayı 2  Ocak-Haziran 2012  (ISSN: 2147-4877, E-ISSN: 2459-0614)
Abidin Çalışkan, Emrullah Acar, Yılmaz Kaya

NO Makale Adı
1356026641 Palm Print Recognition with GLCM based k-NN Classifier Model

Knearest neighboralgorithmis an algorithmthat solvesthe problem
ofclassification. The classification is to analyse texture features of a new image and
to assign them into a predefined class. The important thingis that the features of
each class shoul be determined previously.In this work, the Hong Kong Polytechnic
University Palmprint Database image is used. After the preprocessing of hand images;
the region of interest, namely sub-image of palm print is extracted from the original
images. The texture features are extracted by using theGreyLevelCo-Occurrence
Matrix (GLCM). Then, GLCM based texture features vectors are applied to k-Nearest
Neighbor(k-NN) method as inputs for classification. Finally, the best performance is
observed as % 91.4 in the k=1 structure.