On a taxonomy of ear features

In recent years, the ear modality has attracted attention for personal authentication in both biometrics and forensics. This was motivated by perceived advantages of ear recognition and its wide variety of potential applications. However, to date there is no standard classification scheme (taxonomy) for ear characteristics. This paper organizes ear features in 2D ear images, following a similar organization established for fingerprint and face features. We classify ear characteristic features into three levels, comparing features used by humans and those used by machines for recognition. A set of experiments has been performed to support our proposed classification scheme (taxonomy) based on a mixed set of ear images with multiple scales. The goal of this paper is to provide an insight into the distinctive characteristic features of the ear modality, analyze its anthropometric and morphological specifications, and examine its role in human authentication.