作者还说明了用于体育动作模式识别的匹配隶属函数、模糊神经网络的结构和学习方法。
The membership function of matching movements and the structure and learning methods of fuzzy neural network were also described for recognition of exercise movements.
实验表明这种改进的互功率谱比值法与传统的自功率谱比值法的方法相比,动作模式识别率得到了提高。
The result of experimentation shows that its recognition of movement was much better than the method using traditional Power Spectrum Coefficient.
将其应用于假肢手的控制系统中,通过神经网络进行动作模式识别,共识别了6个手部动作模式,识别成功率在95%以上。
It was applied in the control system of prosthesis hand, the detected signal was recognized by neural network, 6 action patterns can be classified, the success rate is above 95%.
应用推荐