Uncorrelated linear discriminant analysis based on weighted pairwise Fisher criterion

摘要

In this paper, we propose a novel uncorrelated, weighted linear discriminant analysis (UWLDA) method for feature extraction and recognition. The UWLDA first introduces a weighting function to restrain the dominant role of the classes with larger distance and then searches the optimal discriminant vectors under the conjugative orthogonal constrains in the null space of the within-class scatter matrix and its conjugative orthogonal complement space, respectively. As a result, the proposed technique not only derive the optimal and lossless discriminative information, but also guarantee that all extracted features are statistically uncorrelated. Experiments on FERET face database and AR face database are performed to test and evaluate the proposed algorithm. The results demonstrate the effectiveness of UWLDA.