Face Recognition Based on Fusion of 2DLDA and Points of Interest
She Xuebing · 2013
The application of face recognition, two-dimensional linear discriminant analysis ( 2DLDA ) method has good classification results. Density of the interest point detection is based on weighted calculation on the image. The characteristics of the target image with weight is more obvious, and not easy to be influenced by perspective and shape change. Then it is combined with 2DLDA to identify face database. The experimental results show that this method in comparison with the 2DPCA method achieves high recognition rate.