A Two-Stage Perception System Calibration Algorithm for Quadruped Robots Using Factor Graph

The accurate calibration of the perception system is one key prerequisites for quadruped robots to move safely and robustly in challenging environments. Not only does it provide relative transformations sensors w.r.t robot body but also time offset between unsynchronized sensors. This paper presents a two-stage spatio-temporal framework body-eye robots. By attaching complementary sensor, we first estimate sensor frame without complicated modifications mechanical system. Then extrinsic transformation by formulating hand-eye problem while using information from kinematic model encoder measurements. parameters are finally optimized minimizing re-projection error kinematics factor graph formulation with Lie group state representation. experiments demonstrate that proposed algorithm produces precise results which guarantee high-accuracy environmental mapping an RGB-D camera.