Rochester Event Camera Dataset

Rochester Event Camera Dataset 是由苏黎世大学信息学系机器人与感知组发布的全球首个事件相机数据集。该数据集主要用于高速机器人研究，包含事件流、强度图像、惯性测量单元数据以及运动捕捉系统提供的真实轨迹数据。数据集以文本文件和二进制文件格式发布，支持姿态估计、视觉里程计和SLAM等算法的开发与评估。

The Rochester Event Camera Dataset is the world's first event camera dataset released by the Robotics and Perception Group, Department of Informatics, University of Zurich. This dataset is primarily designed for high-speed robotics research, containing event streams, intensity images, inertial measurement unit (IMU) data, and ground-truth trajectory data provided by motion capture systems. The dataset is distributed in both text and binary file formats, supporting the development and evaluation of algorithms such as pose estimation, visual odometry, and SLAM.

搜集汇总

背景与挑战

背景概述

该数据集是世界上首个用于高速机器人技术的事件相机数据集，包含事件数据、强度图像、惯性测量和运动捕捉系统的地面真值。事件相机具有高速、低延迟和高动态范围的优势，适用于姿态估计、视觉里程计和SLAM等任务，数据以文本和二进制（rosbag）两种格式提供。