›› 2016, Vol. 29 ›› Issue (7): 78-.

• 论文 • 上一篇    下一篇

基于Android平台的实时人脸检测与跟踪

于世强,应捷   

  1. (上海理工大学 光电信息与计算机工程学院,上海 200093)
  • 出版日期:2016-07-15 发布日期:2016-07-15
  • 作者简介:于世强(1990-),男,硕士研究生。研究方向:数字图像处理。应捷(1973-),女,副教授,硕士生导师。研究方向:数字图像处理。

Real-time Face Detection and Tracking Based on Android Platform

YU Shiqiang,YING Jie   

  1. (School of Optical-Electrical and Computer Engineering, Shanghai University of Shanghai for Science and Techology, Shanghai 200093, China)
  • Online:2016-07-15 Published:2016-07-15

摘要:

针对Android智能手机自带人脸检测功能效率低、错误率高的问题,提出了一种将OpenCV移植到Android平台的方法,在运行Android系统的嵌入式平台中使用改进的AdaBoost算法,并结合OpenCV库来实现实时人脸检测与跟踪。实验取得了高达9505%的人脸检测准确率和5013 ms的平均检测速率,在保证检测速度的同时比Android自带的人脸检测更具高效性和实用性。

关键词: Android, OpenCV, 人脸检测, AdaBoost算法

Abstract:

Aiming at the low efficiency and high error rate from Android smartphone original face detection system, this paper proposes a method of transplanting OpenCV(Open Source Computer Vision Library) to Android platform, using improved AdaBoost algorithm combined with OpenCV Library to realize real-time face detection and tracking on embedded Android platform. The experiment achieved 9505% detection accuracy rate and 5013 ms average detection speed, it is more effective and practical than Android original face detection system at almost the same detection speed.

Key words: Android, OpenCV, face detection, AdaBoost algorithm

中图分类号: 

  • TP391.41