Independent component analysis human face recognition method based on multi- scale total variation based quotient image

Abstract

The invention discloses a face recognition method by an independent component analysis based on a multi-scale total variational derivative image, which belongs to the face recognition technical field; the method is as follows: a contrast gradient is strengthened; TV-L<1> is used to carry out scale decomposition to a face image to obtain a large-scale image comprising a skeleton contour and muscle information and a small-scale image comprising the details of mouth, eyes and nose; quotient balance is carried out to the small-scale image to obtain the feature of unchanged illumination; feature fusion technology is selected to fuse the features of large scale and unchanged illumination into a new face image; Gabor is used to analyze and extract the features of the new face image in a specific scale and direction to generate a multi-scale Gabor face; the eigenvectors of all the samples are extracted by an information maximization independent component analysis algorithm; the similarity of the eigenvectors of the face which is to be treated with recognizing is calculated by the eigenvectors of the known face; according to the similarity, the eigenvectors are sorted to acquire a final recognition result. The face recognition method achieves high recognition rate and strong robustness to illumination, expression, make-up and other external interference.

Description

technical field [0001] The invention relates to an independent component analysis face recognition method based on a multi-scale overall variation quotient image, and belongs to the technical field of face recognition. Background technique [0002] Illumination conditions are the most important factor affecting the imaging effect of face images. In recent years, researchers have proposed a series of models based on the idea of ​​Retinex for face recognition under complex illumination conditions that do not require prior knowledge of light source characteristics. Wang et al. (IEEE CVPR 2004) proposed a self-quotient image model. Chen et al. (IEEE CVPR 2005) use TV-L 1 model processing images, a proposed method using TV-L 1 Illumination equalization face recognition preprocessing model combined with scale decomposition model and quotient balance model (TVQI). In order to cope with external interference factors such as expression, posture, occlusion, and age in complex backg...

AI Technical Summary

Problems solved by technology

When it is applied to a large-scale face database, its recognition rate is not very ideal, because it is impossible to distinguish a large number of face samples only with limited small-scale information.

At the same time, the TVQI model cannot improve the face recognition rate in other complex backgrounds, that is, it is less robust to external interference factors such as expression, posture, occlusion, age, etc., and sometimes even degrades traditional face recognition methods in these cases. recognition rate

Therefore, when using the TVQI model for real-time face recognition in complex backgrounds, the recognition effect is not ideal

Although the Gabor-based ICA algorithm shows good robustness to external interference factors such as expression, posture, occlusion, and age in the process of face recognition, its robustness against illumination changes is weak.

Images