Face feature extraction method in combination with position information

Abstract

The invention discloses a face feature extraction method in combination with position information, belongs to the technical field of face recognition, and particularly relates to a method for improving face recognition accuracy. According to the invention, on the basis of a face feature extraction method, the calculation dimension of the face position is increased, and clustering processing is carried out on the face according to the position; then abnormal feature data is removed through similarity calculation in the ethnic group; and finally, high-quality face feature data is screened out. By means of the method, high-quality face features can be extracted even under the conditions that the image noise is large, the face part is shielded, the face does not directly face the camera, and the face is far away from the camera, so that the face recognition accuracy is improved.

Description

technical field [0001] The invention belongs to the technical field of face recognition, in particular to a method for improving the accuracy of face recognition. Background technique [0002] Face recognition is a biometric technology that automatically recognizes people based on their facial features (such as statistical or geometric features, etc.). The current general solution is: first detect the face in the image and extract the feature data, then match the extracted face feature data with the feature template stored in the feature database to obtain a similarity, and finally compare the similarity Compared with the preset threshold, if the similarity is greater than the threshold, the recognition is successful. It can be seen that whether high-quality face feature data can be extracted directly affects the accuracy of face recognition. [0003] Face features are usually divided into visual features, pixel statistical features, face image transformation coefficient f...

AI Technical Summary

Problems solved by technology

However, the extraction of face features has relatively high requirements on the environment, background, and light sensitivity. The quality of the face image largely determines the quality of the feature data, especially when the face is partially occluded or the face is not facing the camera. (side face), the face is far away from the camera (small face), the picture is blurred, and the pixels are insufficient, etc., the above method cannot be used to extract high-quality face feature data

Images