摘要
该文提出了一种简单有效的基于人体骨骼化模型的步态识别方法。首先,对输入的步态序列自动进行背景初始化;然后分割图像中运动人体的侧面影像,并进一步细化为人体的骨骼化模型;从模型中提取人体的静态参数(如身高、步幅)以及动态参数(如运动过程中关节点的位置、肢体角度);最后,应用标准的模式分类技术对个体的身份做出识别。实验结果表明,此方法通过提取可靠的步态特征,降低了数据处理的代价,而且得到了较为良好的识别性能。
This article describes a simple and effective gait recognition method based on a skeletal model of the body.At first,the background is initialized automatically in the gait sequence.Secondly,body silhouette is segmented from the image and that is converted into a skeletal model afterwards.And then authors extract body's static and dynamic parameters such as height,stride,the position of the joint and the angle of the body etc.Finally,people can recognize different individual using all of these gait signatures by the pattern classification technology.The utility of the proposed method is illustrated using indoor video sequences in the experiments.And a good identification performance is got.
出处
《计算机工程与应用》
CSCD
北大核心
2005年第9期88-92,共5页
Computer Engineering and Applications
基金
北京市自然科学基金项目(编号:4031004)
北京市教委科技发展项目(编号:km200310005006)
关键词
生物特征识别
步态识别
骨骼化模型
主元分析
特征步态
biometrics,gait recognition,skeletal model,Principal Component Analysis(PCA),eigengait