An Accurate Volumetric Error Modeling Method for A Stereo Vision System Based on Error Decoupling

Stereo vision is a widely used measurement technology, while accuracy is one of its most critical properties. Currently, modeling the volumetric error of the stereo vision system is a way to improve measurement accuracy. This paper proposes an accurate volumetric error modeling method for sub-meter stereo vision systems. An analytical expression of the volumetric error considering the effect of random error is developed at first. To accurately evaluate the measurement error of stereo vision systems, a platform consisting of a laser tracker and a robot is established, and an error evaluation method is proposed. Since the random error is coupled with the measurement error and impacts the accuracy of the volumetric error model, we propose an error decoupling method to separate the random error from the measurement error by only a single measurement. Finally, the accuracy of the proposed method is verified through simulations and experiments.