Hand-eye calibration for robot measuring system based on adaptive differential evolution

Abstract

A platform composed of a line-structured-light vision scanner and an industrial robot is established for 3D surface digitalization. A mathematic model for measuring system is constructed by using homogenous coordinate transform principle and a rough estimation of hand-eye matrix is proposed through measuring a fixed point at different robot hand poses. An adaptive differential evolution, in which the mutation is determined dynamically, is presented for improving accuracy of hand-eye calibration. Test results of real experiments are reported and analyzed.