计算机应用 ›› 2015, Vol. 35 ›› Issue (8): 2321-2326.DOI: 10.11772/j.issn.1001-9081.2015.08.2321

• 虚拟现实与数字媒体 • 上一篇    下一篇

基于三维模型的Android手机端人脸姿态实时估计系统

王海鹏1, 王正良2, 许威威1, 范然1   

  1. 1. 杭州师范大学 杭州国际服务工程学院, 杭州 311121;
    2. 浙江省科技信息研究院 信息资源中心, 杭州 311121
  • 收稿日期:2015-01-28 修回日期:2015-03-28 出版日期:2015-08-10 发布日期:2015-08-14
  • 通讯作者: 许威威(1975-),男,安徽绩溪人,教授,博士,主要研究方向:计算机图形图像处理,weiwei.xu.g@gmail.com
  • 作者简介:王海鹏(1989-),男,山东东营人,硕士研究生,主要研究方向:计算机视觉、计算机图形; 王正良(1978-),男,浙江台州人,馆员,主要研究方向:计算机视觉; 范然(1984-),男,山东济南人,博士,主要研究方向:计算机图形学、三维几何处理。
  • 基金资助:

    国家自然科学基金资助项目(61322204,61272392)。

Real-time face pose estimation system based on 3D face model on Android mobile platform

WANG Haipeng1, WANG Zhengliang2, XU Weiwei1, FAN Ran1   

  1. 1. Institute of Service Engineering, Hangzhou Normal University, Hangzhou Zhejiang 311121, China;
    2. Information Resource Center, Institute of Scientific and Technical Information of Zhejiang Province, Hangzhou Zhejiang 311121, China
  • Received:2015-01-28 Revised:2015-03-28 Online:2015-08-10 Published:2015-08-14

摘要:

针对人脸姿态估计对系统性能要求高、在手机上运行无法满足实时性要求等问题,实现了一种Android手机端的人脸姿态实时估计系统。首先,由摄像头获得一幅正面和一幅偏移一定角度的人脸图像,利用从运动中构建结构(SfM)算法建立简单三维人脸模型;然后,提取实时人脸图像中与三维人脸模型相互对应的特征点,基于缩放正投影位姿估计(POSIT)算法估计人脸姿态角度;最后将三维人脸模型通过开放图形开发库(OpenGL)实时显示在手机屏幕上。实验结果表明,实时视频中检测人脸姿态并显示的速度可以达到20 frame/s,接近计算机端的基于仿射对应的三维人脸姿态估计算法,而且针对大量图片序列的检测可以达到50 frame/s,能够满足Android手机端的性能和检测人脸姿态的实时性要求。

关键词: 人脸姿态, 从运动中构建结构算法, 显示形状回归, 基于缩放正投影位姿估计, 随机抽样一致算法, Android, 增强现实

Abstract:

Concerning that the high performance requirement of face pose estimation system which could not run on mobile phone in real time, a real-time face pose estimation system was realized for Android mobile phone terminals. First of all, one positive face image and one face image with a certain offset angle were obtained by the camera for establishing a simple 3D face model by Structure from Motion (SfM) algorithm. Secondly, the system extracted corresponding feature points from the real-time face image to 3D face model. The 3D face pose parameters were got by POSIT (Pose from Orthography and Scaling with ITeration) algorithm. At last, the 3D face model was displayed on Android mobile terminals in real-time using OpenGL (Open Graphics Library). The experimental results showed that the speed of detecting and displaying the face pose was up to 20 frame/s in the real-time video, which is close to 3D face pose estimation algorithm based on the affine correspondance on computer terminals; and the speed of detecting a large number of image sequences reached 50 frame/s. The results indicate that the system can satisfy the performance requirement for Android mobile phone terminals and real-time requirement of detecting the face pose.

Key words: face pose, Structure from Motion (SfM) algorithm, explicit shape regression, Pose from Orthography and Scaling with ITeration (POSIT), Random Sampling Consensus (RANSAC), Android, augmented reality

中图分类号: