A Review on Techniques for Ear Biometrics

Autor: Michelle Alva, Anuradha Srinivasaraghavan, Kavita Sonawane
Rok vydání: 2019
Předmět:
Zdroj: 2019 IEEE International Conference on Electrical, Computer and Communication Technologies (ICECCT).
DOI: 10.1109/icecct.2019.8869450
Popis: In today’s digital world, security is a major concern. The contribution of biometrics in security has gained momentum due to its various advantages. Ear biometrics is one such domain which uses ear as the primary feature for person recognition. The ear is a promising biometric due to its various unique features. This paper presents a brief study of ear biometrics research focusing on three core phases, namely preprocessing, feature extraction and authentication. Depending on the input ear image quality, various image enhancement techniques are applied by researchers to improve image contents, and the first phase deals with these techniques. The second phase comprises feature extraction techniques experimented upon by various researchers, which are contributing towards extracting the unique characteristics of the image, with the aim to achieve the desired accuracy. The third phase deals with various distance measures and classifiers used for feature vector comparison and decision making for final authentication. This review work also discusses the various performance evaluation parameters and different datasets used by researchers in their experimentation. This paper presents a detailed review of all contributions, highlighting their advantages, disadvantages and constraints, followed by suggested new directions in the same area.
Databáze: OpenAIRE