Design of Experimental Teaching Platform for SLAM System in Low-light and Weak-texture Environment
In order to improve the quality of professional robot experimental teaching and the quality of talent cultivation,this paper proposes a robust visual-inertial fusion SLAM algorithm tailored for low-light and weak-texture environment,and designs a software-based experimental teaching platform for SLAM system in the environment.The proposed approach achieves robust 4DOF state estimation and high-precision real-time localization,and its effectiveness is validated through experiments on publicly available datasets.The experimental results demonstrate that compared to existing visual-inertial SLAM algorithms,the proposed algorithm exhibits higher accuracy and robustness in structured weak-texture scenes,effectively reduces the cumulative errors of the robot system and ensuring map consistency.By utilizing the proposed SLAM system teaching platform in experimental instruction,the innovative and practical capabilities of undergraduate students majoring in robotics are enhanced when dealing with real-world engineering challenges.
NETL
NSTL
万方数据
维普
