基于步态能量图像和2维主成分分析的步态识别方法

Gait Recognition Based on Gait Energy Image and Two Dimensional Principal Component Analysis

为了快速有效地进行步态识别，针对步态能量图像能够表征步态信息和2维主成分分析能快速降维的特点，提出了一种基于步态能量图像和2维主成分分析的步态识别方法。该方法首先应用背景减除法分割出人体轮廓；然后通过人体宽高比的相关信号确定运动周期，再对二值周期序列进行步态能量图像（GEI）合成；最后运用行列相结合的2维主成分分析（（2D）2PCA）方法与加权的2维主成分分析（W（2D）2PCA）方法提取特征主向量，并采用最近邻分类器进行分类。实验结果表明，该步态识别方法可以有效降低前期处理对分类识别的影响，而且对于3种不同行走状态的CASIA数据库中多个视角下拍摄的步态图像可取得很好的识别效果。

In order to carry on the gait recognition fast and effectively, aiming at the characteristics that gait energy image can view the information of gait and two dimension principal component analysis can reduce the dimensions quickly, a novel gait recognition based on gait energy image and two dimension principal component analysis is proposed in this paper. Firstly, the body silhouette extraction is achieved by background subtraction. Secondly, a gait cycle is obtained with the correlated signal of the ratio of width and height of human body. Gait energy image is applied on the binary image sequence to construct the feature vector. Finally, （2D）2PCA and W（2D）2pCA is used to reduce into a low dimension space. The nearest-neighbor classifier is adopted to distinguish the difference. This gait recognition method can decrease the influence of the early preprocess effectively, and our experimental results demonstrate that the method is effective and has achieved a good recognition effect on CASIA gait database including three different multi-views.