Colour-feature dual discriminating correlation analysis for face recognition

How to effectively utilise the colour image information and extract useful features is the key to colour face recognition. In this study, the authors propose a novel colour face recognition approach named colour-feature dual discriminating correlation analysis, which incorporates correlation metric into the discriminant analysis technique, and realises colour-feature discriminating correlation analysis not only within each colour component but also between different components. The public face recognition grand challenge version 2 database is employed as the test data. Experimental results illustrate that the proposed approach outperforms several representative colour face recognition methods.