A comparative study of feature metrics for classification of human passport photos


ASLAN S. , TUNALI T., CINSDIKICI M.

IEEE 15th Signal Processing and Communications Applications Conference, Eskişehir, Turkey, 11 - 13 June 2007, pp.981-984 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Doi Number: 10.1109/siu.2007.4298721
  • City: Eskişehir
  • Country: Turkey
  • Page Numbers: pp.981-984

Abstract

In this study, for a multimedia indexing system that uses a Hierarchical Cellular Tree (HCT), effects of metrics that are obtained from the features extracted by using Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA) are studied for a set of passport size photographs and classification performance is reported. Results obtained reveal that LDA has better performance than that of PCA for HCT.