Hybrid Voice Controller for Intelligent Wheelchair and Rehabilitation Robot Using Voice Recognition and Embedded Technologies

Abstract

The use of intelligent wheelchair and rehabilitation robots has increased rapidly in the recent years owing to a growing number of patients experiencing paralysis, quadriplegia, amputation, and geriatric conditions. In this paper, the design and development of a powerful voice control system is proposed. It includes three modes of operational voice-recognition algorithms. Two sophisticated voice-recognition modules are used to achieve this goal. The system supports speaker dependent (SD) and speaker independent (SI) voice processing. Two voice-recognition algorithms are used, dynamic time warping (DTW) and Hidden Markov Model (HMM), to ensure the maximum voice-recognition accuracy and reduce voice-recognition errors. The system is validated in different noise environments to verify the performance of the system with low and high noise and to evaluate the feasibility of using the system in these environments. Three popular languages (English, German, and Chinese) were used by the system to verify performance with different pronunciations.