Space Target Simulation Dataset in Space-based Detection System

The observation satellite is set to be in a high Earth orbit, observing targets in low Earth orbit.The dataset includes target image sequences under three different illumination backgrounds: strong light glare, non-uniform strong light, and no illumination. Each illumination background contains five sets of target image sequences with different observation axis directions and varying signal-to-noise ratios (SNRs).

The simulated time step for each sequence is 5 seconds, with a total of 1,000 frames per sequence. The image resolution is 500×500 pixels, and the images are stored in PNG format.Additionally, each sequence includes a trace folder, which contains:trace.pngtarget_pos.jsonsnr.txt每个序列还包含一个 GroundTruth 文件夹，用于存储 Ground Truth 目标图像。数据目录结构 -glare_spot：强光眩光背景下的目标数据   --snr1: First set of target image sequences     ---GroundTruth: Ground truth target image sequences       ----0-999.png: Ground truth images     ---trace: Target information       ----snr.txt: Average SNR of the image sequence       ----target_pos.json: Target center position information       ----trace.png: Target trajectory image     ---0-999.png: Simulated target images   --snr2   ...   --snr5 -no_uniform: Target data under non-uniform strong light background   --snr1   ...   --snr5 -no_light: Target data under no illumination conditions   --snr1   ...   --snr5 该数据集提供了不同光照条件下空间目标的真实模拟，支持空间目标检测算法的研发。