Human Identification based on Ear Recognition

Biometrics is the identification of humans based on their characteristics or traits. This paper presents the study of investigating the ear as an effective biometric trait in order to identify humans. The study involves using 2D images of ears and processing them to determine who the individual is. Various feature extraction techniques were explored. The study also presents many methods of classification. Principal Component Analysis (PCA), Local Binary Patterns (LBP), and spatial histograms were used for feature extraction, with spatial histograms yielding the highest success rates. For classification, Support Vector Machine (SVM) and K-nearest neighbour classifiers were used. The highest success rate achieved by the classification was 88%.