Space Target Optical Imaging Simulation in Space-based Passive Optical Detection System

This dataset comprises deep space target observation data generated by the proposed simulation algorithm under space-based passive optical long-distance detection scenarios. It includes three types of illumination backgrounds: strong light spot background, non-uniform strong light background, and no illumination background. For each illumination background, multiple sets of target motion sequence images with different signal-to-noise ratios under various observation conditions were generated. Each set contains 3,000 images with a resolution of 500×500 pixels. The JSON files store the ground truth of target positions for each dataset, and the true trajectories of the targets have also been generated.

本数据集包含由所提出的仿真算法在天基无源光学远距离探测场景下生成的深空目标观测数据。其涵盖三类照明背景：强光光斑背景、非均匀强光背景以及无照明背景。针对每一类照明背景，本数据集生成了多组不同观测条件下、信噪比各异的目标运动序列图像；每组包含3000张分辨率为500×500像素的图像。JSON格式文件存储了各数据集对应的目标位置真值（ground truth），同时也生成了目标的真实运动轨迹。

搜集汇总

背景与挑战

背景概述

该数据集模拟空间被动光学远程探测系统中的深空目标观测，包含强光点、非均匀强光和无光照三种光照背景，每种背景生成多组不同信噪比的目标运动序列图像（每组3,000张500×500像素），并提供目标位置真值和轨迹的JSON文件。