Pose Error Analysis Model Based on Binocular Vision for Rigid-Body

Abstract

In order to satisfy the orientationmeasuring requirements of rigid-body such as work piece, cutting tool inindustry and medical instrumentsfor invasive surgery, this paper presents a binocularvision detection technique based on spatial position information of markers toextract rigid-body pose information and analyzes the pose accuracy of rigid-bodyusing the principalcomponent analysis(PCA) and the least square method(LSM) when spatial position error of markers exist. The simulation experiment demonstrates themaximum angle error of orientation is about 0.59 degree when the position error ofmarkers satisfy the Gaussian distribution with the meanis zero and thestandard deviation is 0~3mm. The experimental results verify this method canrobustly solve the orientation of rigid body using the position information ofmarkers with position errors, and it provides a theoretical and experimentalbasis for orientation measurement of rigid body.

http://dx.doi.org/10.11591/telkomnika.v12i12.6136