摘要
本文主要讲述一种基于多人视频的人脸识别考勤方法,主要应用了人脸检4测,图片拼接以及图片清晰度检测等技术。该方法可以从视频中提取图片,对图片进行人脸检测及图片处理,交给人脸识别服务器对比识别完成签到,相比较于传统的人脸识别考勤机,实现了高并发、过程透明化、防早退等优点,使得该方法可以应用于学校以及各大场景。
This paper mainly describes a face recognition attendance method based on multi-person video, which mainly applies face detection, image mosaic and image sharpness detection technology. This method can extract pictures from videos, detect and process the pictures, and send them to the face recognition server for checking in. Compared with the traditional face recognition attendance machine, this method achieves the advantages of high concurrency, process transparency and early retirement prevention. This method can be applied to schools and major scenarios.
作者
黄志成
周富肯
张杰
贝梓健
廖永发
林焕杜
Huang Zhicheng;Zhou Fuken;Zhang Jie;Bei Zijian;Liao Yongfa;Lin Huandu(Neusoft Institute Guangdong,Foshan Guangdong,528200)
出处
《电子测试》
2019年第8期43-45,共3页
Electronic Test
关键词
人脸识别
基于视频的人脸识别
清晰度算法
face recognition
video-based face recognition
sharpness algorithm