郑庆登,刘蓉,刘佩丽,陈洪波.基于活动摄像头的人脸定位跟踪系统[J].电子测量技术,2016,39(4):62-65
基于活动摄像头的人脸定位跟踪系统
Face detection and tracking system design based on rotatable camera
  
DOI:
中文关键词:  摄像头  人脸检测  Adaboost  OpenCV
英文关键词:camera  face detection  Adaboost  OpenCV
基金项目:国家自然科学基金(81460273)、广西科技攻关计划项目(桂科攻1348020 10)、广西自然科学基金(2013GXNSFA019325)、广西大学生创新项目(201410595102,20141059101)
作者单位
郑庆登 桂林电子科技大学 生命与环境科学学院桂林541004 
刘蓉 桂林电子科技大学 生命与环境科学学院桂林541004 
刘佩丽 桂林电子科技大学 生命与环境科学学院桂林541004 
陈洪波 桂林电子科技大学 生命与环境科学学院桂林541004 
AuthorInstitution
Zheng Qingdeng School of Life and Environmental Sciences, Guilin University of Electronic Technology,Guilin 541004, China 
Liu Rong School of Life and Environmental Sciences, Guilin University of Electronic Technology,Guilin 541004, China 
Liu Peili School of Life and Environmental Sciences, Guilin University of Electronic Technology,Guilin 541004, China 
Chen Hongbo School of Life and Environmental Sciences, Guilin University of Electronic Technology,Guilin 541004, China 
摘要点击次数: 527
全文下载次数: 463
中文摘要:
      针对固定摄像头无法自动跟踪拍摄运动目标的缺陷,设计了基于舵机的活动摄像系统,以实现活动人脸的定位与跟踪。在adaboost人脸检测算法的基础上,结合计算机视觉类库OpenCV与图形界面开发框架Qt在VS2010开发环境下,设计并实现了人脸检测 定位 跟踪系统。测试结果表明,在明亮环境条件下识别准确率达到100%,在昏暗条件下准确率为95%,对于人脸的部分遮挡以及上、下、左、右侧脸都能被检测到。本文方法可广泛应用于安防等领域。
英文摘要:
      For the restrictions that the fixed camera cannot automatically track the moving target, a rotatable camera system was designed based on the steering engine, in order to position and track the moving face. The Adaboost algorithm was used to detect and track the face. The face detection positioning– tracking system was designed under the development environment of Visual studio 2010 with the computer vision class library OpenCV and the graphical interface development framework Qt. The test results show that the accuracy rate for tracing face is 100% in the bright environment, and 95% in the dim environment. And it is also effective to tract the partially occluded face. This system can also be applied for security and other fields.
查看全文  查看/发表评论  下载PDF阅读器