基于多权值神经网络的静态手势字母识别的研究

Recognizing for hand alphabet gestures based on multi-weighted neurons

提出了一种基于多权值神经网络模型的静态手势语识别方法。应用手势字母图像圆周极径序列的傅立叶频谱信息来提取特征,再结合多权值神经网络的训练算法与识别算法,实现静态手势字母的识别,并取得了很好的识别效果。

With the development of human computer interaction technology, the research of understanding for human＇s gesture has been an important task for discussion. This paper proposed a new method for recognizing of hand alphabet gestures, which based on multi-weighted neurons. The method applied Fourier-descriptors for feature extraction and combined with the classification algorithm of multi-weighted neurons for recognizing, the results showed that the method performed well.