Pose estimation based on PnP algorithm for the racket of table tennis robot

Abstract

An approach was presented to detect the racket position and pose. The racket region is first extracted by the segmentation technique with the color information. The improved LSD algorithm was investigated to detect the line segments in the racket region image and then straight lines were extracted by cluster and the image coordinates of the vertices of the rectangle were obtained. Finally, the 3D coordinates of the racket face center and the racket pose are estimated via PnP positioning Method. The experimental results verify the effectiveness of the proposed method.