Scholarlink

<p>As SLAM technology has evolved from the geometric level to the object level, active SLAM (ASLAM) has also adopted a new goal: improving the ability to observe objects. The robot's observation views and its motion have a significant impact on object-level SLAM. However, current ASLAM mostly focuses on low-dimensional environmental features such as points, lines, and planes. This paper proposes a new observation trajectory optimization method based on the ellipsoid model and the camera field of view, aiming to enhance object observability. We integrate view planning and motion planning into a unified optimization module, and construct an optimized factor graph based on probabilistic inference. Leveraging the camera field of view and the object ellipsoid model, three key factors are introduced into the graph: object completeness observation factor, self-observation prevention factor, and camera motion smoothness factor. Our method guarantees globally optimal observation trajectories considering both object observation effectiveness and robot motion performance. Finally, we implement this method in a complete object-level ASLAM system, and evaluate it in both simulated and real-world environments. Experimental results demonstrate that our method significantly improves object modeling accuracy, mapping efficiency, and localization precision. The code for this work is open source at <a href="https://github.com/TINY-KE/OTO_ASLAM.git">https://github.com/TINY-KE/OTO_ASLAM.git</a>.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Robot observation trajectory optimization for active object SLAM with ellipsoid model and camera field of view

  • 期刊论文

  • 发表日期：

    2025年12月5日

查看全文
  • Jiadong Zhang,
  • Wei Wang

摘要

As SLAM technology has evolved from the geometric level to the object level, active SLAM (ASLAM) has also adopted a new goal: improving the ability to observe objects. The robot's observation views and its motion have a significant impact on object-level SLAM. However, current ASLAM mostly focuses on low-dimensional environmental features such as points, lines, and planes. This paper proposes a new observation trajectory optimization method based on the ellipsoid model and the camera field of view, aiming to enhance object observability. We integrate view planning and motion planning into a unified optimization module, and construct an optimized factor graph based on probabilistic inference. Leveraging the camera field of view and the object ellipsoid model, three key factors are introduced into the graph: object completeness observation factor, self-observation prevention factor, and camera motion smoothness factor. Our method guarantees globally optimal observation trajectories considering both object observation effectiveness and robot motion performance. Finally, we implement this method in a complete object-level ASLAM system, and evaluate it in both simulated and real-world environments. Experimental results demonstrate that our method significantly improves object modeling accuracy, mapping efficiency, and localization precision. The code for this work is open source at https://github.com/TINY-KE/OTO_ASLAM.git.