Abstract: Currently, the method of theoretical analysis and static verification is mainly used to evaluate the accuracy of visual measurement data in dynamic visual measurement tasks, which is inadequate for assessing the impact of error factors on measurement results during dynamic visual measurement processes. To address tihe limitation, this study designed and constructed an image dynamic pose measurement accuracy evaluation system. Based on the visual measurement principle and typical target motion patterns, the system comprehensively analyzed the error influence factors, planned and designed diverse experimental scenarios. Leveraging the mobile robot's advantages, including controllable motion, real-time dynamic pose acquisition and high-precision repeated positioning/posing, a ground simulation of target motion was conducted. By comparing the benchmark data generated by the mobile robot with the pose data obtained through visual measurement, error tracing and quantitative analysis were achieved, which provided effective technical support for the optimization design, feasibility demonstration and accuracy evaluation of specific taskthe implementation plan.