DiTer++
收藏资源简介:
DiTer++数据集是由仁荷大学和韩国机械材料研究所创建的,旨在为多机器人SLAM在多会话环境中的应用提供多样化的地形和多模态数据。该数据集包含多个机器人(Agent-A和Agent-B)在不同时间段(白天和夜晚)采集的数据,涵盖了多种传感器配置,如RGB相机、LiDAR和热成像仪。数据集的创建过程包括生成高精度的先验地图,并通过扫描到地图匹配技术提取每个机器人的真实轨迹。DiTer++数据集主要应用于多机器人SLAM任务,旨在解决大规模环境中的定位与建图问题,特别是在动态和光照变化频繁的环境中。
The DiTer++ dataset was developed by Inha University and the Korea Institute of Machinery and Materials, aiming to provide diversified terrain and multimodal data for the application of multi-robot SLAM in multi-session environments. This dataset includes data collected by multiple robots (Agent-A and Agent-B) across different time periods (daytime and nighttime), covering various sensor configurations such as RGB cameras, LiDAR, and thermal imagers. The dataset creation process involves generating high-precision prior maps and extracting the true trajectories of each robot through scan-to-map matching technology. The DiTer++ dataset is primarily used for multi-robot SLAM tasks, targeting the resolution of localization and mapping problems in large-scale environments, particularly those with dynamic changes and frequent illumination variations.




