rUNSWift SPL Field Segmentation Dataset

rUNSWift SPL Field Segmentation Dataset是由新南威尔士大学计算机科学与工程学院的团队rUNSWift创建的，旨在解决RoboCup SPL比赛中机器人视觉系统的足球场分割问题。该数据集包含20个视频，记录了Nao V5/V6人形机器人在不同环境条件下的行为，每个视频包含连续的高分辨率帧及其对应的标签。数据集的创建过程涉及在多种光照和行为设置下收集数据，并通过特定的工具进行标注。该数据集主要应用于开发和验证机器人视觉算法，特别是在复杂光照条件下的足球场分割。

rUNSWift SPL Field Segmentation Dataset was created by the rUNSWift team from the School of Computer Science and Engineering, University of New South Wales, with the goal of addressing the soccer field segmentation problem for robotic vision systems in RoboCup SPL competitions. This dataset includes 20 videos recording the operational activities of Nao V5/V6 humanoid robots under various environmental conditions. Each video contains consecutive high-resolution frames and their corresponding pixel-wise annotation masks. The dataset was collected under diverse lighting conditions and operational settings, with annotations generated using specialized tools. It is primarily utilized for developing and validating robotic vision algorithms, especially for soccer field segmentation under complex lighting environments.