[1]李思璇,孙 兵.人脸对齐处理对面部色泽变化的影响[J].医学信息,2019,(19):4-7.[doi:10.3969/j.issn.1006-1959.2019.19.002]
 LI Si-xuan,SUN Bing.Analyzing the Influence of Face Alignment Processing on Facial Color Change[J].Medical Information,2019,(19):4-7.[doi:10.3969/j.issn.1006-1959.2019.19.002]
点击复制

人脸对齐处理对面部色泽变化的影响()
分享到:

医学信息[ISSN:1006-1959/CN:61-1278/R]

卷:
期数:
2019年19期
页码:
4-7
栏目:
医学信息学
出版日期:
2019-10-01

文章信息/Info

Title:
Analyzing the Influence of Face Alignment Processing on Facial Color Change
文章编号:
1006-1959(2019)19-0004-04
作者:
李思璇1孙 兵2
(北京航空航天大学生物与医学工程学院1,电子信息工程学院2,北京 100191)
Author(s):
LI Si-xuan1SUN Bing2
(College of Biological and Medical Engineering1,College of Electronic and Information Engineering2,Beijing University of Aeronautics and Astronautics,Beijing 100191,China)
关键词:
人脸对齐面部色泽变化健康警示相关函数
Keywords:
Face alignmentFacial color changeHealth warningCorrelation function
分类号:
TP391
DOI:
10.3969/j.issn.1006-1959.2019.19.002
文献标志码:
B
摘要:
对人脸识别考勤系统所获取的图像序列进行色泽变化分析,可有效获取员工的健康信息。本文利用基于Haar-like特征的人脸识别算法自动确定左眼、右眼、鼻子和嘴巴特征点位置,分别采用左右眼两点对齐法、左右眼+鼻子三点对齐法、左右眼+嘴巴三点对齐法,对图像序列进行配准处理以得到标准化图像序列,对标准化图像序列的各色彩分量进行相关函数分析,研究人脸图像序列对齐处理方法对面部色泽变化的影响。探讨能够有效减小图像序列对齐处理引起的面部色泽变化的方法,更好地保留员工自身健康变化引起的面部色泽变化,提供更为准确的健康警示信息。
Abstract:
The color image change analysis of the image sequence obtained by the face recognition attendance system can effectively obtain the employee's health information. In this paper, the face recognition algorithm based on Haar-like feature is used to automatically determine the position of the left eye, right eye, nose and mouth feature points, respectively, using the left and right eye two-point alignment method, left and right eye + nose three-point alignment method, left and right eyes + mouth three point alignment method is used to register the image sequence to obtain a standardized image sequence, and the correlation function analysis is performed on each color component of the normalized image sequence, and the face image sequence alignment processing method affects the color change of the face. To explore ways to effectively reduce facial color changes caused by image sequence alignment processing, to better preserve facial color changes caused by employees' own health changes, and to provide more accurate health warning information.

参考文献/References:

[1]张建树.人脸对齐与图像分类关键算法研究[D].浙江师范大学,2018. [2]王文峰,李大湘,王栋,等.人脸识别原理与实战[M].北京:电子工业出版社,2018:70-84. [3]黄海波,肖子曾,向忠军,等.中医药治疗亚健康的进展[J].中医药导报,2015,21(5):50-52. [4]孙琦,李新霞,武建,等.TCMI智能中诊平台的设计与开发[J].医学信息,2018,31(7):7-9. [5]王天芳,李灿东,朱文锋.中医四诊操作规范专家共识[J].中华中医药杂志,2018,33(1):185-192. [6]梁玉梅.基于面象特征的中医体质自动辨识系统研究[D].北京工业大学,2016. [7]张红凯,李福凤.中医面诊信息采集与识别方法研究进展[J].世界科学技术-中医药现代化,2015,17(2):400-404. [8]蔡轶珩,吕慧娟,郭松,等.中医望诊图像信息标准量化与显示复现[J].北京工业大学学报,2014,40(3):466-472. [9]王祉,张红凯,李福凤,等.中医面诊信息计算机识别方法研究及临床应用概述[J].中华中医药学刊,2014(8):1882-1885. [10]王晓亚,封筠.人脸识别考勤技术研究[J].河北省科学院学报,2014,31(2):54-59.

更新日期/Last Update: 2019-10-01