基于可变区域特征和SVM的步态识别研究

Gait recognition research based on region variance feature and SVM

根据人体随步态变化不一样的特点,提出了一种基于人体可变区域分割的步态识别方法。首先,应用背景差方法分割出运动人体轮廓,然后将人体分为多个可变区域,并通过计算获取特征向量。最后对得到的特征量采用SVM进行步态的分类和识别。在UCSD和CASIA步态数据库上进行实验,结果表明该方法不但能克服由于获取的特征量过少而造成的信息丢失,还取得了较好的识别性能。

According to the different feature between body and gait variance, a new gait recognition approach based on region variance feature was presented. First, the background subtraction was used to extract body silhouette. Then the dimensional silhouette was divided into several regions and feature vectors were acquired. Finally, gait classification and recognition were performed by support vector machine. The method was applied to two data-sets （UCSD and CASIA）. Experimental results demonstrate that not only can the approach overcome the information lost that results from a few feature vectors but also get satisfying recognition performance.