An improved head pose estimation method for the robotic wheelchair interaction control

Abstract

Concerning the physically handicapped people, using head movement to control the motion of mobile service robot has become a new choice, and it has favored by many consumers and researchers and has high market value and research value. In view of the shortcomings of the traditional head pose estimation algorithm in practical applications, this paper proposes an improved head pose estimation method based on the combination of random forest and iterative closest point (ICP) algorithm for robotic wheelchair interactive control. And the corresponding experiments are conducted to compare the accuracy and time performance of this algorithm with the traditional algorithm.