生活场景语义地图构建及实时更新数据集

本数据集围绕生活支援机器人在室内动态场景下的精细环境感知与长期自主导航问题，通过仿真推演与实机测试相结合的方式构建。数据生产于2023年9月至2025年6月，空间范围为重庆大学自动化学院构建的标准化室内家居模拟环境。数据集核心内容包括基于Cartographer算法构建的室内二维栅格地图、基于YOLOv5模型生成的语义标签与物体检测信息、激光雷达与深度相机采集的多模态传感器数据，以及基于ROS的语义地图融合与实时更新算法代码包。数据经过严格的传感器校准、时间同步校验与地图一致性验证，确保了其完整性与准确性。本数据集为验证语义SLAM系统的建图精度、语义信息关联有效性以及导航系统在动态环境中的鲁棒性提供了关键数据支撑，对推动服务机器人环境认知与智能导航技术的研发与重用具有重要价值。

This dataset is constructed by combining simulation deduction and physical robot testing, focusing on the issues of precise environmental perception and long-term autonomous navigation of life-support robots in indoor dynamic scenarios. The data was generated from September 2023 to June 2025, and the spatial scope covers the standardized indoor home simulation environment built by the School of Automation, Chongqing University. The core contents of the dataset include 2D indoor grid maps constructed based on the Cartographer algorithm, semantic labels and object detection information generated by the YOLOv5 model, multimodal sensor data collected by LiDAR and depth cameras, as well as ROS-based code packages for semantic map fusion and real-time update algorithms. All data has undergone strict sensor calibration, time synchronization verification and map consistency validation to ensure its integrity and accuracy. This dataset provides key data support for verifying the mapping accuracy of semantic SLAM systems, the effectiveness of semantic information association, and the robustness of navigation systems in dynamic environments, and has important value for promoting the research, development and reuse of service robots' environmental cognition and intelligent navigation technologies.

搜集汇总

背景与挑战

背景概述

该数据集专注于室内生活场景，为生活支援机器人提供语义地图构建和实时更新的关键数据。它包含多模态传感器数据、语义标签和算法代码，用于验证语义SLAM系统的精度和导航鲁棒性，支持服务机器人环境认知技术的研发。