Aiming at the problem that it is difficult to accurately identify movement of the finger only using electromyographic（EMG） signal of the arm surface

a method combining the pose signal of finger joint with the surface EMG signal for identifying the posture of the finger is proposed. Use MYO bracelet synchronously acquire 8 sets of EMG signals of the arm

at the same time

use gesture sensors to capture the three-dimensional angle information of movement of finger. Moving average energy method is used to detect the active segment based on the collected raw EMG signal signals

and extract related signal in the process of execution

and extract the features of the EMG signal and angle signal respectively by means of mean absolute value（MAV） and standard deviation（SD）. The extracted eigenvalues are used as the input of the support vector machine（SVM） multi-class classifier to identify gesture actions. The experimental results show that the fusion of finger angle signals can effectively improve the recognition rate of finger movements

the overall recognition rate is 99. 3 %.