Statistical Movement Classification based on Gram-Schmidt Transform

In this work, a technique for automatic classification of finger motion signals was introduced by using Gram-Schmidt orthonormalization. Because the motion signal has nonlinear and non-stationary processing characteristics, the Gram-Schmidt orthonormalization method is used to analyze the motion signal. By batch processing the high correlation between variables, the classification results are obtained. At the same time, combined with the idea of partial least squares, a new identity measure is proposed to select the best parameters. Finally, SVM is used to classify the movement types of the research objects. According to the test sample records of two finger motion signals (thumb and pinky), the average accuracy is 96.65%. The average accuracy of data activation (thumb and pinky) was 100%. Experiments show that the operation speed of Gram-Schmidt classification method is significantly improved, and it has better interpretability and prediction ability.