Face recognition attendance checking method based on width learning

Abstract

The invention provides a face recognition attendance checking method based on width learning. The face recognition attendance checking method based on width learning comprises the steps that a certainnumber of to-be-checked face images are collected and integrated into a database; preprocessing the face image; carrying out face detection to position an image area where the face features are located; carrying out static feature extraction to construct a face feature library, and dividing the face feature library into a training set and a test set; performing training of a width learning network by using the training set, constructing a width learning basic model, performing test comparison by using the test set and the trained width learning network, and performing test comparison to obtain a misjudged face image; collecting misjudgment face images during test comparison as a new feature vector set, and performing incremental learning on the width learning basic model to optimize the width learning basic model; and performing face recognition on the to-be-checked person, matching a face recognition result with the identity information in the database, making a corresponding check-in record, and storing the check-in record information in the database.

Description

technical field [0001] The invention relates to the technical field of face recognition, in particular to a face recognition attendance method based on breadth learning. Background technique [0002] The algorithms used in the existing face recognition college classroom attendance system are roughly divided into recognition algorithms based on geometric features, recognition algorithms based on principal component analysis, and recognition algorithms based on deep learning including but not limited to CNN. Among them, the CNN-based system implementation process is mainly divided into image acquisition module, face detection module, deep learning training module, feature extraction module, matching recognition module, etc. [0003] The existing system has achieved very good results in large-scale face data processing, but for a university student, the number of students is thousands or even tens of thousands, so I want to design a system with a higher recognition rate A deep...

AI Technical Summary

Problems solved by technology

[0004] The technical problem to be solved by the present invention is to provide a face re

Images