写字楼室内场景定位与建图标准数据集

主要面向自主机器人与人工智能研究、高质量多传感器SLAM数据集需求建设，基于自主研发的一体式多传感器平台采集产生，基于双目彩色相机、双目灰度相机、激光雷达、轮速计、IMU等传感器进行高时空精度的同步采集，采集场景为上海交通大学电信群楼，场景特点为存在较多且密集的桌、椅、门、窗、柱等室内典型部件。存在弱纹理(白墙)、弱结构(长走廊)等挑战性因素等。

Targeting the requirements of autonomous robotics and artificial intelligence research for high-quality multi-sensor SLAM datasets, this dataset was collected via a self-developed integrated multi-sensor platform. Synchronized data acquisition with high spatio-temporal precision was performed using multiple sensors including binocular color cameras, binocular grayscale cameras, LiDAR, wheel odometers, and IMUs. The data collection was conducted in the Telecommunication Building of Shanghai Jiao Tong University, which features dense indoor typical components such as tables, chairs, doors, windows, and pillars. Additionally, the scene contains challenging scenarios including low-texture regions (e.g., white walls) and low-structure environments (e.g., long corridors).

搜集汇总

背景与挑战

背景概述

该数据集是一个面向自主机器人与人工智能研究的多传感器SLAM数据集，基于双目彩色相机、双目灰度相机、激光雷达、轮速计、IMU等传感器在上海交通大学电信群楼采集，包含弱纹理和弱结构等挑战性场景，数据量为25.67GB。