基于频率感知记忆增强的三域特征学习用于移动红外小目标检测
摘要
作为目标检测的一个子领域，移动红外小目标检测由于目标尺寸微小且与背景的对比度低，面临着显著的挑战。目前的方法主要仅依赖于时空域提取的特征，而频率域几乎未受到关注，尽管在图像处理中已广泛应用。为扩展特征来源域并增强特征表示，我们提出了一种新的三域策略（Tridos），并在红外小目标检测中引入频率感知记忆增强以用于时空域。在该方案中，一个局部-全局频率感知模块通过傅里叶变换有效地分离并增强频率特征。受人类视觉系统的启发，我们的记忆增强设计用于捕获红外目标在视频帧之间的空间关系。此外，我们进一步通过差分学习和残差增强对其进行编码，以捕获时序动态运动特征。 Additionally, we further design a residual compensation to reconcile possible cross-domain feature mismatches. To our best knowledge, proposed Tridos is the first work to explore infrared target feature learning comprehensively in spatio-temporal-frequency domains. The extensive experiments on three datasets (i.e., DAUB, ITSDT-15K and IRDST) validate that our triple-domain infrared feature learning scheme could often be obviously superior to state-of-the-art ones. Source codes are available at https://github.com/UESTC-nnLab/Tridos.
引用
@article{arxiv.2406.06949,
title = {Triple-domain Feature Learning with Frequency-aware Memory Enhancement for Moving Infrared Small Target Detection},
author = {Weiwei Duan and Luping Ji and Shengjia Chen and Sicheng Zhu and Mao Ye},
journal= {arXiv preprint arXiv:2406.06949},
year = {2024}
}
备注
This paper has accepted IEEE TGRS