Abstract: This paper built a unimodal biometric system that is suitable for most individual modalities, e.g., face and gait. We first preprocess each raw datum into unit and then obtain a new biometric data set. Then we determine the intrinsic low-dimensional subspace of preprocessed data by local topology structure preserving projections (LTSPP). Finally we perform the classification in the determined subspace using the intra-class distance sum. In the proposed system, LTSPP is a novel subspace algorithm. Compared with other subspace methods, LTSPP possesses more discriminant abilities and is more suitable for biometric recognition. In addition, both preprocessing each raw datum into unit and performing the classification using the intra-class distance sum are helpful to improve the recognition rates. The experimental results demonstrate the effectiveness of our unimodal biometric system.

Key words: unimodal biometric system, subspace selection, local topology structure preserving projections (LTSPP), intraclass distance sum