Age estimation method and system based on multi-scale linear differential textural features

Abstract

The invention discloses an age estimation method and system based on multi-scale linear differential textural features. The method includes the steps of: first, using Haar-like features and an Adaboost classifier algorithm to detect a face in a picture to be detected, and cutting out a face region; then extracting textural features based on multi-scale linear differential features from the face region that is cut out, using a decision-making tree to perform feature selection, and reducing feature vector dimensionality; and finally training an SVR model in a feature space after the feature vector dimensionality is reduced, and using the SVR model to perform age prediction of the face picture. Since RBF kernel-based nonlinear support vector regression is used as a classifier model to perform age estimation, the age estimation is relatively accurate, and in addition, the age estimation is performed through the face picture, thereby enabling human-computer interaction to be relatively humanized and safe.

Description

technical field [0001] The invention relates to the technical field of biological identification, in particular to an age estimation method and system based on multi-scale linear difference texture features. Background technique [0002] Face pictures contain rich personal information, including identity, age, gender, and race, which are widely used in the field of human-computer interaction. With the development of e-commerce and the popularization of various mobile devices, user's age information plays an increasingly important role in human-computer interaction. Age estimation based on face pictures has broad application prospects in human-computer interaction, such as security monitoring, user management in e-commerce, website access control, image and video retrieval, and more humanized human-computer interaction functions. [0003] Relying on the biometric feature of the face for age estimation does not require active cooperation from the user, so the operation is hig...

AI Technical Summary

Problems solved by technology

[0004] However, in the existing human-computer interaction process, the judgment of the user's age has problems such as difficult identification or large judgment errors.

Images