论文部分内容阅读
提出了基于线性判别分析(LDA)和支持向量机(SVM)的步态识别方法。应用背景差方法分割出运动人体轮廓,对人体轮廓进行水平和垂直投影,将垂直和水平投影向量合成为一维步态数据向量。通过线性判别分析对步态数据进行特征提取和压缩,对得到的识别量应用支持向量机进行步态的分类和识别。应用上述方法在CMU步态数据库上进行了实验,实验结果表明本文所提的步态识别方法具有较高的识别性能。
A gait recognition method based on Linear Discriminant Analysis (LDA) and Support Vector Machine (SVM) is proposed. The background of the moving human body is segmented by using the background subtraction method. The contour of the human body is projected horizontally and vertically, and the vertical and horizontal projection vectors are synthesized into one-dimensional gait data vector. The features of gait data were extracted and compressed by linear discriminant analysis, and the classification and identification of gait were implemented by using SVM. Experiments on CMU gait database using the above method show that the gait recognition method proposed in this paper has high recognition performance.