| 摘要: |
| 目的 利用移动终端的传感设备， 将广告投放对象精准化。方法 采用基于肤色修正的Haar特征人脸检测算法、 粒子滤波和Mean-Shift相结合的人脸追踪算法以及Gabor特征性别判别和年龄估计算法， 设计基于云计算的硬件架构， 展望匿名视频， 分析系统发展趋势。结果 在云端实现了对广告关注人数、 关注时间及关注者年龄、 性别的实时统计分析并出具了相关报告。结论 对在不同时刻、 不同地点投放广告从而获取最大效益的实验提供了有力的数据支撑。 |
| 关键词: 广告注意力 人脸检测 性别判别 年龄估计 云计算 |
| DOI： |
| 分类号:TB472 |
| 基金项目:国家科技支撑计划项目 （2012BAH85F02） |
|
| Digital Signage Effect Based on Face Cloud Date |
|
SHEN Hao1, CUI Cheng2, WANG Xin2, JI Tie1
|
|
1.Hunan University， Changsha 410082， China;2.Harbin Institute of Technology Shenzhen Graduate School， Shenzhen 518055， China
|
| Abstract: |
| Using sensing devices of mobile terminal，putting the advertisement accuracy will be the most important research direction.Skin fixed Haar feature algorithm for face detection， particle filter with Mean-Shift algorithm for face tracking， Gabor feature for gender discrimination and age prediction are used.It also designs a hardware structure based on Cloud computing，prospects the anonymous video and analyzes the development trend. These make the real-time statistical analysis of the number of advertisement followers，attention time，ages and genders come true.These statistics will provide a powerful basis of when， where and which advertisement to play can get the most effective results.v |
| Key words: advertising attention face detection gender discrimination age prediction cloud computing |
渝公网安备 50010702501716号