Ava Tahmasebi - Young Researchers Club, NajafabadBranch, Islamic Azad University,Isfahan, Iran
Hossein Pourghassem - Department of ElectricalEngineering, Najafabad Branch,Islamic Azad University, Isfahan,Iran
Homayoun Mahdavi-Nasab - Department of ElectricalEngineering, Najafabad Branch,Islamic Azad University, Isfahan,Iran

Biometric recognition systems are inquired to solvesecurity problems in networked society recently. Uniqueness ofear shape and its robustness in spite of increasing age hasattracted researcher’s attention. In this paper, an appropriatebiometric system is presented based on local-Gabor featureswhich effectively extract appropriate features bring alongdimensionality reduction. Subsequently the KNN classifier isapplied. Experimental results show the effectiveness of theproposed method which brings both higher recognitionperformance and higher speed benefits

Ear recognition, Gabor filter, local features, KNNclassifier.I. INTRODUCTIONS