遇见数据集

Tracking boats in the strait of hormuz using NASIR imagery and higher dimensions

收藏
Zenodo2026-05-14 更新2026-05-26 收录
官方服务:

资源简介:

NISAR-Radar Dataset represents a high-fidelity topological reconstruction of maritime and terrestrial backscatter derived from Level-1 Range-Doppler Single Look Complex (RSLC) radar geometry. By mapping the intensity of radar returns to the vertical displacement ($z$-axis), the engine transforms 2D satellite observations into a physical Manifold where man-made structures—specifically steel-hulled freighters and urban infrastructure—manifest as peaks of energy. This 12-dimensional vector space $(t, x, y, z, R, G, B, N_x, N_y, N_z, d, \theta)$ allows for the precise mathematical isolation of non-natural anomalies. In this dataset, the Theta Angle ($\theta$) serves as a primary diagnostic for Ridges, identifying the sharp geometric discontinuities where high-density reflectors meet the fluid or organic baseline of the ocean and land. Through this transformation, the dataset moves beyond imagery, providing a kinematic record of structural energy signatures moving across the planetary surface in 12-day orbital intervals.

NISAR雷达数据集是基于一级距离多普勒单视复(RSLC)雷达几何数据生成的高保真海洋与陆地后向散射拓扑重构成果。通过将雷达回波强度映射至垂直位移(z轴)维度,该引擎将二维卫星观测数据转换为物理流形(Manifold)。在此流形中,人工构筑物——具体为钢壳货轮与城市基础设施——将以能量峰值的形式显现。 该数据集构建的12维向量空间$(t, x, y, z, R, G, B, N_x, N_y, N_z, d, heta)$支持通过精确数学方法分离非自然异常目标。在本数据集中,θ角(Theta Angle)可作为识别脊状特征的核心诊断指标,用于定位高密度反射体与海洋、陆地的流体或有机基底相交处的尖锐几何不连续区域。通过该转换流程,本数据集突破了传统影像数据的局限,可提供以12天轨道周期为间隔的行星表面结构能量特征运动学记录。

提供机构:
Zenodo
创建时间:
2026-05-13
二维码
社区交流群
二维码
科研交流群
商业服务