Decision level fusion of hybrid local features for face recognition

In this paper, we propose a feature extraction method based on decision level fusion of local features for simple yet robust face recognition. The origin face is first divided into smaller regions from which local binary pattern (LBP) histogram sequences are extracted and concatenated into a global feature representation. In addition, statistical texture information is also exploited to fuse the results with LBP features at decision level in order to enhance the performance. The recognition is evaluated using different similarity measures on public face databases. Preliminary results demonstrate that the proposed algorithm is efficient and suitable for real time face recognition application.