Publication

Abstract : Over the last decade, biometrics have lately been receiving researchers' considerable attention. Traditional means of human identification are PIN's, login IDs, passwords, and identification cards. The major issues with these traditional identification systems are: hard to remember, easy to lose and can be stolen. On the other hand, biometrics characteristics are universal, unique, permanent and measurable. Human ear is a major source of person identification; therefore, it is widely used for security in public places, forensic investigations, e-banking, etc. This paper critically illustrates numerous approaches for human ear identification. As a sequel to a survey on the problems in the concerned fields, it also provides different solutions to overcome the problems. The different methods adopted by researchers in the past decade for ear detection are Harr-, shape-, wavelet-, subregion-and template-based. Major dataset is provided with analyses which have been used by the researchers in different reported articles. This review paper provides major guidelines to the upcoming researchers interested in ear biometric field.