让小卫星实时检测洪水，仅用极少存储就保持高精度
Towards Onboard Continuous Change Detection for Floods
- 用历史注入机制让Transformer记住过往影像，存图量减少99%以上
- 在STTORM-CD数据集上，精度接近双时相基线模型
- 可在纳米卫星硬件上实现43帧每秒，适合实时灾害监测
通过持续卫星观测进行自然灾害监测，需在严格运行约束下处理多时相数据。本文针对灾害管理关键任务——洪水检测，提出一种可在小型卫星上运行的机载变化检测系统，满足内存与计算限制。我们提出面向Transformer模型的历史注入机制（HiT），在保留前序观测历史上下文的同时，将数据存储量压缩至原始图像大小的1%以下。在STTORM-CD洪水数据集上的测试表明，基于普里蒂维-小型基础模型的HiT-Prithvi模型，在检测精度上与双时相基线相当。该模型在代表性的纳米卫星硬件Jetson Orin Nano上达到43 FPS。本工作构建了卫星连续监测自然灾害的实用框架，支持无需地面处理基础设施的实时灾害评估。架构及模型权重已公开于https://github.com/zaitra/HiT-change-detection。
原文摘要 · Abstract (English)
Natural disaster monitoring through continuous satellite observation requires processing multi-temporal data under strict operational constraints. This paper addresses flood detection, a critical application for hazard management, by developing an onboard change detection system that operates within the memory and computational limits of small satellites. We propose History Injection mechanism for Transformer models (HiT), that maintains historical context from previous observations while reducing data storage by over 99\% of original image size. Moreover, testing on the STTORM-CD flood dataset confirms that the HiT mechanism within the Prithvi-tiny foundation model maintains detection accuracy compared to the bi-temporal baseline. The proposed HiT-Prithvi model achieved 43 FPS on Jetson Orin Nano, a representative onboard hardware used in nanosats. This work establishes a practical framework for satellite-based continuous monitoring of natural disasters, supporting real-time hazard assessment without dependency on ground-based processing infrastructure. Architecture as well as model checkpoints is available at https://github.com/zaitra/HiT-change-detection .
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。