Thalmic Labs: EMG Pattern Recognition

拟议的研究将集中于开发能够处理多个通道的表面检测肌电(SEMG)信号的信号处理和模式识别算法，以提取与执行动态手势和手臂手势相关的信息。所开发的算法的输出旨在用于控制各种计算机软件应用。表面肌电信号将使用应用于前臂的现有8个电极和专用数据采集系统来采集。表面肌电信号将被处理以提取与各种手势或手或手臂动作相关的特征。这些特征将被输入到一系列模式识别算法中，以确定当前正在执行的手势。开发的算法将是正在设计和开发的整体硬件和软件设备的组成部分，以实现与计算机软件应用程序的轻松和无线交互。最终目标是为计算机、移动设备或机器人等机器创造一种新颖的接口设备，从而创造一种基于自由空间中自然手势和运动的新用户输入方法。

The proposed research will focus on the development of signal processing and pattern recognition algorithms capable of processing mulitple channels of surface-detected EMG (sEMG) signals to extract information related to the execution of dynamic hand and arm gestures. The output of the developed algorithms are intended to be used to control various computer software appliactions. The sEMG signals are to be acquired using an existing band of 8 electrodes applied to the forearm and a dedicated data acquisition system. The sEMG signals will be processed to extract features related to various gestures or hand or arm movements. These features will be input to a series of pattern recognition algorithms devloped to determine which gesture is currently being executed. The devloped algorithms will be an integral part of an overall hardware and software device being designed and developed to allow effortless and wireless interaction to computer software applications. The ultimate goal is to create a novel interface device for machines such as computers, mobile devices, or robotics, thereby creating a new method of user input based on natural gestures and movements in free-space.