基于WPD和（2D）2PCA的步态识别方法
Gait Recognition Using WPD and (2D)2PCA
-
摘要: 为了提高步态识别率,在步态能量图(gait energy image,GEI)基础上,提出了基于小波包分解(waveletpacket decomposition,WPD)和完全主成分分析(two-directional two-dimensional principal component analysis,(2D)2PCA)的步态识别方法.该方法采用基于人体轮廓的GEI来解决步态数据量过大的问题,并采用WPD和(2D)2PCA进行步态特征提取,解决了已有基于小波变换的步态识别方法中高频分量丢失或维数过高问题.在NLPR步态数据库上对该方法进行了评测,并与经典方法进行了比较.实验结果表明:该方法具有更高的识别率和视角变化的鲁棒性.Abstract: To improve the gait recognition rate,a gait recognition method based on wavelet packet decomposition(WPD) and two-directional two-dimensional principal component analysis((2D) 2PCA) was proposed.In the method,gait energy image(GEI) of the body silhouette was firstly adopted to solve the problem of huge gait data.And then,WPD and(2D) 2PCA were used to extract features to solve the problem existing in the gait recognition method based on wavelet transform at present: high frequency component loss or high dimension problem.The Experiment evaluation was conducted on NLPR gait database and compared with classical methods.Result shows that the proposed method has a higher recognition rate and is more robust to the change of view.
下载: