VBR: A Vision Benchmark in Rome
收藏资源简介:
VBR数据集是由罗马大学计算机、控制与管理工程系创建,旨在为视觉和感知研究提供一个全面的基准。该数据集包含RGB数据、3D点云、IMU和GPS数据,覆盖了罗马的多个特征区域,总轨迹超过40公里,记录时间近4小时。数据集通过手持和车载数据收集方式,模拟各种机器人平台,如四足机器人、四旋翼飞行器和自动驾驶车辆。数据集创建过程中,使用最新设备并实施了有效的传感器校准和时间同步程序。VBR数据集的应用领域包括自主机器人和计算机视觉,特别是针对视觉测距和SLAM的研究,旨在解决环境多样性、运动模式和传感器频率等问题。
The VBR Dataset was developed by the Department of Computer, Control and Management Engineering of Sapienza University of Rome, aiming to provide a comprehensive benchmark for visual and perceptual research. It contains RGB data, 3D point clouds, IMU and GPS data, covering multiple characteristic regions in Rome. The total trajectory exceeds 40 kilometers, with a total recording duration of nearly 4 hours. Data was collected via handheld and in-vehicle acquisition methods to simulate deployment on various robotic platforms, including quadruped robots, quadrotor UAVs, and autonomous vehicles. During the dataset construction process, state-of-the-art equipment was utilized, and effective sensor calibration and time synchronization procedures were implemented. The application domains of the VBR Dataset include autonomous robotics and computer vision, particularly research on visual odometry and SLAM, and it is designed to address issues such as environmental diversity, motion patterns, and sensor frequency limitations.

- 1VBR: A Vision Benchmark in Rome罗马大学计算机、控制与管理工程系“Antonio Ruberti” · 2024年



