Abstract Flood verification and mitigation based on ground survey has been impeded by many factors. For instance, damage assessment exclusively by onsite expedition could be delayed by restricted access, especially during heavy flooding. To devise appropriate mitigation measures, resources required to thoroughly investigate the incidents in response to disaster could be overwhelming. Moreover, inadequate equipment and non-standardized flood reporting process could also undermine the maneuver. As a consequent, rescue mission was inevitably deferred until the reported incidents had been confirmed. To address these issues, this paper presents an internetworking system for assisting flooding disaster mitigation. Its main contributions were two folds. Firstly, its data intelligent was driven by convergence of both user and official reports of flood incidents and those automatically detected from remotely sensed images. Secondly, the reliability of reported information was ensured by two-factor verification, i.e. by using a state-of-the-art deep learning strategy and by official investigations. The verified incidents were then presented in augmented graphical reports, which were practical in public inquiry and official uses in devising efficient and cost-effective disaster mitigation measures. The developed system was deployed in different cities, most affected by recent floods. Numerical and subjective assessments herein demonstrate the merits of the system.

中文翻译：

---

基于遥感和移动GIS的互联互通防洪减灾系统

摘要 基于地面调查的防洪减灾受到诸多因素的阻碍。例如，完全由现场探险进行的损害评估可能会因进入受限而延迟，尤其是在洪水泛滥期间。为了制定适当的缓解措施，彻底调查事件以应对灾难所需的资源可能是压倒性的。此外，设备不足和不规范的洪水报告流程也可能会破坏机动。因此，救援任务不可避免地被推迟，直到所报告的事件得到确认。为了解决这些问题，本文提出了一种用于协助减轻洪水灾害的网络互联系统。它的主要贡献有两个方面。首先，它的数据智能化是由用户和官方的洪水事件报告以及从遥感图像中自动检测到的数据的融合驱动的。其次，通过使用最先进的深度学习策略和官方调查的双因素验证来确保报告信息的可靠性。经核实的事件随后以增强的图形报告形式呈现，这在公共调查和官方用途中非常实用，可用于制定高效且具有成本效益的减灾措施。开发的系统部署在不同城市，受近期洪水影响最大。此处的数值和主观评估证明了该系统的优点。报告信息的可靠性由两因素验证确保，即通过使用最先进的深度学习策略和官方调查。经核实的事件随后以增强的图形报告形式呈现，这在公共调查和官方用途中非常实用，可用于制定高效且具有成本效益的减灾措施。开发的系统部署在不同的城市，受近期洪水的影响最大。此处的数值和主观评估证明了该系统的优点。报告信息的可靠性由两因素验证确保，即通过使用最先进的深度学习策略和官方调查。经核实的事件随后以增强的图形报告形式呈现，这在公共调查和官方用途中非常实用，可用于制定高效且具有成本效益的减灾措施。开发的系统部署在不同城市，受近期洪水影响最大。此处的数值和主观评估证明了该系统的优点。开发的系统部署在不同城市，受近期洪水影响最大。此处的数值和主观评估证明了该系统的优点。开发的系统部署在不同城市，受近期洪水影响最大。此处的数值和主观评估证明了该系统的优点。