Dataset for “A Hausdorff-Guided Deep Learning Approach for Monitoring the Motion of Rotating Arctic Ice Floes”

Data for manuscript submitted to GIScience & Remote Sensing entitled “A Hausdorff-Guided Deep Learning Approach for Monitoring the Motion of Rotating Arctic Ice Floes.”

The dataset contains training data used for the deep learning framework developed in the study. The data consist of multi-temporal Arctic sea ice images used to train and evaluate the proposed Hausdorff-guided method for monitoring the motion and rotation of Arctic ice floes.

Each data folder includes the input ice floe images used for model training in the manuscript. The dataset supports the development and validation of the deep learning model used to detect feature points and estimate the motion of rotating ice floes under complex Arctic marginal ice zone conditions.

本数据集为提交至《地理信息科学与遥感》（GIScience & Remote Sensing）期刊的论文配套数据，该论文题为《面向北极旋转浮冰运动监测的豪斯多夫（Hausdorff）引导深度学习方法》。本数据集包含本研究中开发的深度学习框架所用的训练数据，其采用多时序北极海冰影像，用于训练与评估本文提出的豪斯多夫引导式北极浮冰运动与旋转监测方法。每个数据文件夹均收录论文中用于模型训练的输入浮冰影像。本数据集可支撑在复杂北极边缘冰区条件下，用于检测特征点并估算旋转浮冰运动的深度学习模型的开发与验证工作。