摘要
在复杂背景下进行头部检测与跟踪是多功能感知的重要交互通道。在对传统头部检测算法进行改进的基础上,提出了一种基于"粒子"滤波器的实时头部跟踪算法。试验结果表明该算法实时性强,且具有良好的鲁棒性,能在头部发生转动、遮挡及多人干扰等复杂环境下进行稳定跟踪,可广泛应用于头部跟踪、图像分割和人脸检测等领域。
Detecting and tracking human head in a complex background is one of the most important multimodal human-machine interfaces. A real time human head-tracking algorithm that is based on particle filter and improved head detecting algorithm is presented. Experimental results show that this algorithm can detect and track human head satisfactorily in various complex conditions such as head turning, sheltering, among multiple persons or in unconstrained backgrounds. This algorithm can be used in head tracking, image segmentation and face recognize.
出处
《计算机工程与设计》
CSCD
北大核心
2005年第2期354-355,368,共3页
Computer Engineering and Design
基金
陕西省科学技术研究发展基金项目(2003k06-G15)。
