摘要
人体运动分析在计算机视觉研究领域中具有广泛的应用前景。人体骨架包含了很多的人体运动信息,且简单明了,所以对人体骨架提取具有很重要的意义。本文提出了一种人体骨架模型的建立方法,该方法在骨架提取方面有较强的去噪和抗干扰能力,有很好的鲁棒性,可提取更加有效的骨架图像,进一步结合人体解剖学的基础,建立了人体骨架模型,为人体特征识别(HID)和运动人体的实时追踪打下了坚实的基础。
The human body movement analysis has the widespread application prospect in the computer vision territory of research. The human body skeleton contained the very many human bodies movement information, also simple was bright, therefore withdrew to the human body skeleton has the very vital significance. This article proposed one human body skeleton model establishment method, this method withdrew the aspect in the skeleton has stronger goes chirp and the antijumming ability, had the very good robustness. It can effective remove the noise, withdraws the more effective skeleton picture, and in the human body anatomy foundation, established the human body skeleton model, (HID) and movement human body real-time tracing built the solid foundation for the human body characteristic recognition.
出处
《微计算机信息》
北大核心
2006年第07S期227-229,共3页
Control & Automation
基金
北京市"现代信息科学与网络技术"重点实验室资助项目(基金项目编号为:TDXX0503)