Scholarlink

Subterranean Simultaneous Localization and Mapping (Sub-T SLAM) is an active and challenging research area that is generally tackled with high-end hardware and multiple sensors. In this paper, we focus on RGB-only visual SLAM in underground environments, specifically in loop closure detection under opposite viewpoints. We propose a solution using Visual Transformers (ViT) features and simple feature matching that shows promising results.

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

Visual SLAM in Underground Environments: Preliminary Results

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Bliman Federico,
  • Monzon Pablo,
  • Llofriu Martin

摘要

Subterranean Simultaneous Localization and Mapping (Sub-T SLAM) is an active and challenging research area that is generally tackled with high-end hardware and multiple sensors. In this paper, we focus on RGB-only visual SLAM in underground environments, specifically in loop closure detection under opposite viewpoints. We propose a solution using Visual Transformers (ViT) features and simple feature matching that shows promising results.