Spotted words recognition system based on Kalman filter and HMM (hidden Markove model) models to control the movement of the manipulator arm

Abstract

The main objective of this paper is to implement a real-time speech recognition module to use it in controlling the movements of a five degree of freedom (FOD) manipulator arm using a Kalman filter as a selector in noisy environment and HHM model to recognise 12 Arabic spotted words. The adopted methodology is based on detecting and spotting vocabulary words within a phrase generated by user. The system recognises the spotted words using Kalman filter to select these spotted words then a robust HMM (hidden Markove model) technique with cepstral coefficients is used to improve the recognition rate. To implement the approach on a real-time application, a “personal computer parallel port interface” was designed to control the movement of a set of stepper motors. The user can control the movements of five degree of freedom (DOF) for a robot arm using a vocal phrase containing spotted words. Key words: Hidden Markove model, kinematics, Kalman filter, recognition, manipulator, robot, degree of freedom, artificial intelligence.