The Rosario Dataset v2
收藏资源简介:
The Rosario Dataset v2是一个多模态数据集,由阿根廷国家科学技术研究理事会(CONICET)的CIFASIS机构收集于大豆田中,包含超过两小时的传感器记录数据,包括立体红外相机、彩色相机、加速度计、陀螺仪、磁力计、GNSS和轮式里程计。该数据集旨在解决农业环境中机器人定位、建图、感知和导航的挑战,支持高级算法的开发和基准测试。数据集包含6个序列,机器人覆盖总距离7.3公里,记录时间为2.2小时,数据涵盖了自然光照变化、运动模糊、粗糙地形和长轨迹等挑战。
The Rosario Dataset v2 is a multimodal dataset collected in soybean fields by the CIFASIS institute under the National Scientific and Technical Research Council of Argentina (CONICET). It contains over two hours of sensor-recorded data, including stereo infrared cameras, color cameras, accelerometers, gyroscopes, magnetometers, GNSS, and wheel odometry. This dataset aims to address challenges in robot localization, mapping, perception and navigation in agricultural environments, and supports the development and benchmarking of advanced algorithms. The dataset consists of 6 sequences, with the robot covering a total distance of 7.3 kilometers over a recording period of 2.2 hours. The data covers challenges such as natural lighting variations, motion blur, rough terrain and long trajectories.




