Multi-pose face recognition using head pose estimation and PCA approach

In this paper, a novel multi-pose face image recognizing method is presented. In this algorithm, the face with multi-pose can be recognized by comparing it with the eigenface which is generated from 3D face model database, view point of the 3D face model is estimated from the face image, the estimation algorithm utilize the facial geometrical feature to estimate head pose parameter which can be applied to the view point of 3D face model. Through the method, the 3D face model can keep the same pose with the real captured face. Finally, a PCA based algorithm is employed to extract the eigenface from the generated exemplar database and input face image. The cosine distance matching method will be used to compare the similarity of face between input one and the generated database. The one which has the maximum similarity can be judged as the positive one of identifying face. In the experiment, we evaluated the efficiency pose estimation algorithm and the recognition approach. We can see the error of the estimation algorithm is near to ±0.05o for frontal face and ±3.9o for near profile face. And the correct recognition rate is close to 96% for frontal face and 72% for near profile face.