Compound Structure Classifier System for Ear Recognition

Ear recognition is a new research area in the computer vision and pattern recognition field. This paper proposes a new ear biometrics system-compound structure classifier system for ear recognition (CSCSER), based on the research of ear recognition with algebraic feature. The system first makes rough classification to the human ears according to their geometric features. Then the algebra features are extracted and used for detailed classification. Finally the results are achieved, which are in accordance with human natural recognition process. The experiments show that the system can achieve high recognition rates and is suitable for complex ear image libraries.