基于模糊支持向量机的步态识别

提出基于模糊支持向量机（FSVM）的步态识别方法，以人体步态的宽度向量作为特征，探讨直接取值法和模糊C均值2种模糊隶属度确定方法对FSVM步态分类效果的影响。实验结果表明，模糊C均值法的识别率均略好于SVM，直接取值法的识别率甚至低于SVM，因此，选取正确的模糊隶属度确定方法是FSVM能否成功应用于步态识别的关键。

Gait Recognition Based on Fuzzy Support Vector Machine

This paper proposes a gait recognition approach based on Fuzzy Support Vector Machine(FSVM). By using the width vector as feature, it discusses the influence of assign value algorithm and fuzzy C-means algorithm on the recognition rate of gait recognition. Experimental results show that compared with Support Vector Machine(SVM), fuzzy C-means algorithm improves the effectiveness of the classification with FSVM, but the recognition rate of the assign value algorithm is lower than with SVM, which indicates that choosing an appropriate fuzzy membership is the key for FSVM to be applied in gait recognition successfully.