基于激光融合与彩色视觉的SLAM陪伴机器人系统设计

Design of SLAM Companion Robot System Based on Laser Fusion and Color Vision

为了提高教育机器人的定位精准度,此次研究对机器人的激光雷达观测模型和彩色相机成像模型进行了研究,提出了四种观测模型误差,并通过融合激光视觉技术对误差进行消除,设计了优化后的机器人SLAM方案。实验结果表明,此次研究提出的融合激光视觉技术的定位误差仅有1.8%,并可以在3.2 s内完成机器人定位。在静态和动态环境中,该技术分别可以实现98.5%和92.6%的导航成功率。且优化后的机器人可以达到92.8%的满意度。故此次研究提出的融合激光视觉技术的SLAM机器人在学前教育领域可以有效地发挥陪伴作用。

In order to improve the positioning accuracy of educational robots,this study investigated the LIDAR observation model and color camera imaging model of robots,proposed four observation model errors,and eliminated the errors by fused laser vision technology,and designed an optimized robot positioning self-navigation scheme.The experimental results show that the fused laser vision technique proposed in this study has a localization error of only 1.8%and can complete robot localization within 3.2s.In both static and dynamic environments,this technology can achieve navigation success rates of 98.5%and 92.6%,respectively.And the optimized robot can achieve a satisfaction rate of 92.8%.Therefore,the SLAM robot proposed in this study that integrates laser vision technology can effectively play a companion role in the field of preschool education.