Design and implementation of human-computer interaction intelligent system based on speech control

Abstract

Introducing the voice control technology into the human-computer interaction of the mobile terminal can inject new vitality into the terminal device, simplify the operation difficulty, provide a brand-new experience for the user, and have a great market prospect. This paper discusses the sampling of speech signal and the method of frame windowing. Based on this, the method of extracting the characteristic coefficient of speech signal and its key parameters are discussed, including the concepts of short-term energy, power, zero-crossing mean in time domain and LPCC and MFCC parameters in the frequency domain. The DTW and HMM algorithms of speech recognition are studied. After describing their respective characteristics and solving methods, they are selected. Meanwhile, an embedded system software architecture was built. In order to verify the feasibility of the control system, the voice program was verified, including endpoint detection and DTW algorithm program. The verification results show that the voice control method of this paper is accurate and efficient for the control of human-computer interaction intelligent system.