一种基于平均步态能量图的身份识别算法

A Human Recognition Scheme Based on Mean Gait Energy Image

提出一种基于步态能量图(GEI)的嵌入式隐马尔可夫模型(e-HMM)身份识别方法。首先通过预处理提取出运动人体的侧面轮廓,根据步态下肢的摆动距离统计出步态周期,得到平均步态能量图。对能量图的各区域进行分析,利用二维离散余弦变换(2D-DCT)将能量图观测块转化为观测向量,实现嵌入式隐马尔可夫模型的训练和身份识别。最后在USF和CASIA步态数据库上对所提出的算法进行实验。实验表明该方法具有较好的识别性能,是一种有效的步态识别方法。

An embedded hidden Markov model（e-HMM） human recognition scheme based on gait energy image（GEI） is proposed.First a preprocess technique is used to segment the moving silhouette from the walking figure.The algorithm obtains the gait quasi-periodicity through analyzing the width information of the lower limbs’ gait contour edge,and the mean GEI is calculated from gait periodic.It makes use of an optimized set of observation vectors obtained from the two dimensional discrete cosine transform（2D-DCT） coefficients of the mean GEI regions.The e-HMM is trained and used for the gait recognition.The proposed algorithm is evaluated on USF and CASIA Gait Database.The experimental result shows that the proposed approach is valid and has encouraging recognition performance.