Recognizing human subjects by their ears has been a field of great interest for a long time. It has gained serious attention in the recent past due to technological advances. Although several surveys on ear biometrics exist in the literature, Abaza et al. expand on existing work through an analysis of more than 50 publications that emerged during the period from 2007 to 2010. A detailed section of the paper is devoted to ear databases used in the literature for testing proposed schemes.
One section of the paper elaborates on the anatomy of the ear and its evolution through the embryonic stages of the human life cycle. Detecting the ear within a target image (generally of a face) is the first and most important step toward building efficient ear recognition systems. The authors provide an in-depth analysis of the existing research on face detection and a thorough review of the existing work on recognizing faces.
The paper also discusses a related maturing field of study: ear-based multimodal biometric systems. The analysis is very helpful in assessing the pros and cons of the multimodal techniques reported in the literature.
Abaza et al.’s presentation is intended for readers with a background in biometric research. For those interested in specific disciplines such as computer vision, algorithms, and signal processing, this paper is certainly worth reading.