废墟室外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，采集场景为上海市杨树浦路凉州路某拆迁小区，场景特点为存在地形复杂、复杂纹理（外观凌乱、辨识力差）、复杂结构（结构混乱）等挑战性因素。This dataset is developed to fulfill the requirements of autonomous robotics and artificial intelligence research, as well as the demand for high-quality multi-sensor SLAM datasets. It was collected using an independently developed integrated multi-sensor platform, with synchronized data of high spatiotemporal precision acquired via multiple sensors including binocular color cameras, binocular grayscale cameras, LiDAR, wheel odometers, and IMUs. The data collection was carried out in a demolished residential neighborhood at the intersection of Yangshupu Road and Liangzhou Road, Shanghai. The scene presents several challenging conditions: complex terrain, complex textures (disordered appearance with poor recognizability), and disorganized complex structures.

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等传感器在上海市杨树浦路凉州路某拆迁小区进行高时空精度的同步采集，场景特点包括地形复杂、复杂纹理和复杂结构。数据集总大小为61.3GB，数据格式包括.bag、.tiff、.txt和.rar。