一种基于稀疏表示的手势识别算法

A gesture recognition algorithm based on sparse representation

针对经典手势识别方法中在旋转变化与偏移情况下识别率不高的问题,提出一种基于稀疏表示的手势识别算法。通过最小二乘法求解手势特征的稀疏表示,利用Sebastien手势库训练构建出稀疏表示手势冗余字典,最后根据残差最小值实现手势识别。实验结果表明:在手势发生旋转变化和偏移的情况下,所提出的基于稀疏表示的手势识别算法识别率高于经典的最近邻分类手势识别算法。

Sparse representation is proposed for the classic gesture recognition method under an angle rotation and offset variation, in which the recognition rate is not high. Sparse representation is solved by the least square meth- od, then the redundant dictionary is gained from the Sebastien training samples, and will be used to sparsely repre- sent the test gestures to classify the gesture images by the residual error minimum classification. The experimental results show that the test signal samples which were classified by sparse representation could be identified with a higher rate than the classic neighbor nearest classification tion of the gesture. , even if there is a certain angle rotation and offset varia-tion of the gesture.