A gender recognition method and device based on face images

Abstract

The present invention provides a gender recognition method and device based on human face images, wherein the gender recognition method based on human face images includes: obtaining gender recognition results of consecutive N frames of human face images, and the gender recognition results are used to represent the identity of the observed person Gender attribute, N is a natural number greater than 1 and less than the preset number of frames; judge whether the gender recognition results of consecutive N frames of human face images are consistent; when the gender recognition results of the continuous N frames of human face images are consistent, the The gender attribute of the observer is determined as the gender attribute represented by the gender recognition result. In this way, the selected number of image frames is smaller than the number of image frames selected by the average method when obtaining consecutive N frames of human face images with consistent gender recognition results, so as to improve the recognition rate. And the stable result can be expressed by the power of the recognition accuracy of a single frame face image under the number of selected image frames, so when the number of selected image frames is small, a higher stable result can be obtained, that is, a relatively stable gender recognition result.

Description

technical field [0001] The invention relates to the technical fields of computer vision and face recognition, in particular to a gender recognition method and device based on a face image. Background technique [0002] The gender recognition method based on face images is a process in which face image information uses computer technology to identify the gender attributes of the observed person, mainly including: face image preprocessing, face gender feature extraction, face gender classification, and recognition result stabilization. step. Among them, face gender feature extraction is the premise of gender classification algorithm. The quality of gender feature extraction directly affects the performance of final gender classification, and stable recognition results will improve user experience. [0003] However, in the video stream, when shooting the same person, due to slight changes in face position, angle, lighting and other conditions, the stability of face recognition...

AI Technical Summary

Problems solved by technology

However, if this averaging method wants to obtain a more stable gender recognition result, it needs to use more frames of face images, which reduces the recognition rate.

Images