복수 해상도 구순문 인식에 관한 연구 = A Study on Lip Print Recognition by Multi-Resolution Architecture

다국어 초록 (Multilingual Abstract)

This paper presents a lip print recognition by pattern kernels for a person identification. A lip print recognition has been developed less than the other human physical attributes of a fingerprint, a voice pattern, a retinal blood/vessel pattern, or a facial recognition. The merit of a lip print recognition by CCTV camera is linked with other recognition systems such as retinal/iris eye and face. A new method with multi-resolution architecture is proposed to recognize a lip print by the pattern kernels. A set of the pattern kernels is a function of some local lip print pattern masks. This function converts the information on a lip print into the digital data. The recognition in the multi-resolution system is more reliable than recognition in the single resolution system. The results show that the proposed algorithm by the multi-resolution architecture can be efficiently realized.