"Aliased Small Target (AST) Dataset"

"To validate the proposed intelligent deformed small target recognition algorithm in aliased images, we construct the DM-ST Dataset (Detector Multiplexing Small Target Dataset). The dataset comprises 18 motion trajectories of small infrared targets, covering both full-field and local field-of-view regions in forward and reverse directions. Images are acquired via detector multiplexing, producing aliasing effects, and are modulated by field-of-view shape coding to reflect geometric distortions induced by cylindrical mirror modulation and off-axis aberrations. Gaussian and salt-and-pepper noise are added in the simulation to emulate realistic detector readout and space environment interference. Each trajectory contains 600 frames at 60 fps, enabling joint trajectory\u2013shape feature analysis and small target recognition performance evaluation. The DM-ST Dataset provides a high-fidelity benchmark for both simulation and experimental validation of the proposed algorithm."

为验证所提出的混叠图像下智能形变小目标识别算法，我们构建了DM-ST数据集（Detector Multiplexing Small Target Dataset）。该数据集包含18条红外小目标运动轨迹，覆盖正、反向拍摄场景下的全视场与局部视场区域。图像通过探测器复用方式采集，会产生混叠效应，并经视场形状编码调制，以反映柱面镜调制与离轴像差所引发的几何畸变。仿真过程中额外加入了高斯噪声与椒盐噪声，用以模拟真实探测器读出噪声与空间环境干扰。每条轨迹以60 fps（帧每秒）的帧率包含600帧图像，可支撑联合轨迹-形状特征分析以及小目标识别性能评估任务。DM-ST数据集可为所提算法的仿真与实验验证提供高保真基准测试集。