摘要
为了提高课堂管理效率,基于OpenCV开源视觉库,开发出了课堂行为采集管理平台。通过人脸录入、识别等技术,实现了无人值守考勤信息自动采集与识别;通过消息通知模块对缺勤数据及时上报辅导员、家长等,做到缺勤信息及时掌握;通过数据管理模块,实现平台业务数据无死角全备份,实现学生的缺勤申诉做到有数据可查;最后优化了LBPH算法,解决了面部遮挡、相似样本不确定性等问题,取得较为理想的识别率。
In order to improve the efficiency of classroom management,a classroom behavior collection and management platform has been developed based on the OpenCV open-source visual library.Through technologies such as facial input and recognition,automatic collection and recognition of unmanned attendance information has been achieved.Absenteeism data is reported to counselors,parents,etc.in a timely manner through the message notification module,ensuring timely mastery of absenteeism information.Through the data management module,full backup of platform business data is achieved without dead corners to ensure that there is data available for students'absenteeism complaints.Finally,the LBPH algorithm was optimized to address issues such as facial occlusion and uncertainty of similar samples,achieving ideal recognition rates.
作者
孙颖
龚雄涛
SUN Ying;GONG Xiong tao(Hubei Polytechnic Institute,Xiaogan,Hubei 432100,China)
出处
《湖北职业技术学院学报》
2023年第3期89-94,78,共7页
Journal of Hubei Polytechnic Institute
基金
2019年教育部高校产学研创新基金资助项目“基于人工智能的高职教学管理系统研究”(项目编号:2019ITA03017)
2020年教育部首批国家级职业教育教师教学创新团队课题研究项目“‘1+X证书制度’在人工智能技术与应用共同体专业人才培养试点探索与实践应用”(项目编号:SJ2020040101)的阶段性研究成果。
关键词
人脸识别
无人值守
学情分析
语音播报
face recognition
unmanned attendance
learning analytics
voice announcement
