The extraction of key-posture frames from the video of 3-D motion of a human

Abstract

The proposed humanoid robot (HR) with a stereo vision system (SVS) captures a sequence of 3-D motion images of a human, which is faced to the proposed HR. After the inquiry of enough motion sequences, the motion detection via image ratio and background registration, the estimation of five feature points (i.e., head, four tips of two arms and legs) of human body. Based on the silhouettes of human body with five estimated feature points, the key-posture frames from the video of 3-D motion of a human are extracted. They are then applied to imitate the 3-D motion of a human for an HR. Finally, the corresponding experimental results are presented to confirm the usefulness of the proposed method.