Method for establishing facial feature database in video image

Abstract

The invention provides a method for establishing a facial feature database in a video image. On the basis of a locally linear embedding (LLE) algorithm, a face manifold unit in the video image is learned, high-dimensional image space of the image is reduced to low-dimensional image space, and a relationship between an internal low-dimensional structure and high-dimensional observation data is searched; and clustering is carried out on the low-dimensional image space by using a K-means clustering algorithm, so that similar face image data are close, a face of each clustering center is used as a representative face image, a redundant example is removed, and a representative face feature base is established finally. In the LLE algorithm, the embedding dimension of the low-dimensional space is determined according to a Sammon coefficient. According to the invention, facial feature base establishment is realized in the video image; and some representative images are selected from many training images and are used for follow-up algorithm learning. Therefore, while the face identification rate is kept well, the training time and the occupied internal storage are reduced substantially, so that the processing speed increases.

Description

technical field [0001] The invention relates to a face recognition method, in particular to a method for establishing a face feature database in video images, and belongs to the technical field of face recognition of video images. Background technique [0002] The early face recognition methods were all based on the geometric features of the face. The face is composed of eyes, nose, mouth, chin and other parts. The differences in the shape, size and structure of these parts make every face in the world very different. Therefore, these The geometric description of the shape of the components and the mechanical relationship can be used as important features for identification. This type of method is not very sensitive to the illumination and viewing angle of the face image, but it needs to accurately locate the positions of the eyes, nose, mouth, chin and other parts. It should be known that the early face recognition methods based on geometric features were all manually extra...

AI Technical Summary

Problems solved by technology

However, most algorithms ignore the motion information of the face, and only use the spatial structure information. For example, some algorithms first select frames with good quality, and then use the recognition technology of static face images for recognition; while some algorithms for Each frame of image is recognized by a static face recognition method, and then joint decision-making is made on the results of each frame; and some algorithms that combine spatial information and time information have achieved gratifying results. However, how to combine the spatial structure of the face and the dynamics of the face information to identify, so far, there is no relevant literature to elaborate

Images