Face Fusion Recognition Based on Bit-Plane Image

ABSTRACT

Feature extraction is the main technology in the process of human face recognition system which is widely utilized as the mainstream approach in identity verification system. However, the traditional human face feature extraction algorithm of Eigenface has strict restraint on the light in the original human face image, posing as a hindrance in the practical application. Aiming at this problem, this paper proposes a novel method for human face feature extraction based on human face bit-plane image using Eigenface and Fisher's linear discriminant analysis at the same time. The constitution and characteristic of the fusion recognition method is analyzed and the experiments on performance study are proved the superiority of the new fusion recognition method.

Keywords:

Bit-plane image, fisher’s linear discriminant analysis, human face feature extraction, principle component analysis.