A Study on Recognition Algorithm of Ultrasonic Sensing Gesture Based on Improved Hidden Markov Model

Gesture is a convenient means of human-computer interaction. After ultrasonic sensing and extraction of gesture feature sequence, Hidden Markov Model is commonly used to recognize gesture categories. Aiming at the problem that the accuracy of gesture recognition algorithm based on conventional Hidden Markov Model is unsatisfactory, a recognition algorithm of ultrasonic sensing gesture based on improved Hidden Markov Model was proposed in this paper. In this algorithm, state transition probability matrix was improved by Support Vector Machine, and the output probabilities of hidden states in the state sequence were processed by Sigmoid function to optimize the classification performance, so as to improve the accuracy of gesture recognition. In this paper, eight gesture recognition experiments were carried out and the test results showed that the improved algorithm based on Hidden Markov Model optimized by Support Vector Machine could accurately recognize gesture, and the average recognition rate is 94.625 percent, which is 10.35 percent higher than that of conventional Hidden Markov Model. And for each gesture, the recognition rate of the proposed method in this paper was higher than that of based on conventional Hidden Markov Model.