본문영역 바로가기 메인메뉴 바로가기 하단링크 바로가기
정보공개
ENG
검색하기
서울연구원의 연구자료를 검색하세요
사이트맵 열기
모바일 메뉴 열기

카드뉴스

서울연구원에서 만드는 카드뉴스입니다.

Real-Time Deep Learning Analysis to Secure the Golden Time for Flood Response
  • 호13
  • 등록일2025.11.17
  • 조회수783
  • UrbanFlooding
  • FloodMonitoring
  • DeepLearning
  • SmartCitySeoul
  • ClimateResilience
  • RealTimeDetection
  • UrbanSafety
  • AIinCities
  • FloodResponse
  • DisasterTech

서울연구원 card news real time deep learning analysis to secure the golden time for flood response
Flood Risks Are Growing Localized heavy rainfall is becoming more frequent in Seoul due to climate change. Traditional drainage systems alone are no longer enough to prevent urban flooding. A new system that enables real-time detection and rapid response is now essential.
Limits of Traditional Monitoring Flood warning systems have long relied on rainfall data and physical sensors. But these systems are expensive to install and often fail to detect highly localized or unexpected flooding.
이것도 부탁해 AM 10:23 Deep Learning Flood Detection To address this, the Seoul Institute has developed a Real-Time Urban Flood Monitoring System using CCTV footage and deep learning. The technology identifies flooding directly from video—without the need for physical sensors. Real-Time Road Flood Depth Analysis Using Deep Learning Road Flooding Image Collection Flood Depth Analysis Image Data Generation (Level 3 0.49 / Level 4 1.20)
AI Flood-Level Assessment The AI compares water height with vehicle tire height to determine flood depth. It allows fast and accurate assessment without sending staff to the field, greatly improving emergency response efficiency. Flood Depth Level Criteria Level 1: Road Surface ~ 12cm Level 2: 12cm ~ 35cm Level 3: 35cm ~ 60cm Level 4: 60cm ~ 150cm
Tested in Gangnam: It Works A pilot test in Seoul's Gangnam district operated 24/7. The system successfully detected flooding in real time and helped secure the crucial
Cities need more intelligent flood-response strategies as climate risks grow. Seoul's new smart flood-management model shows strong potential. For more details, please refer to the full research report on the Seoul Institute website. Report Development and Application of a Deep Learning-Based Real-Time Urban Flood Monitoring System Sung Eun Kim · Jongrak Baek Card News by Seunghoon Oh, PR & Cooperation Team The Seoul Institute (서울연구원)

[Card News Episode 13] Real-Time Deep Learning Analysis to Secure the Golden Time for Flood Response

As urban flooding becomes more frequent due to climate change, the limitations of traditional flood-response systems have become increasingly evident. In response, the Seoul Institute has developed a Real-Time Urban Flood Monitoring System that uses CCTV footage and deep learning technology. Pilot operations in select areas of Seoul showed that the system is highly effective in securing the crucial “golden time” for flood response. We look forward to a smarter, faster, and more efficient flood-management strategy for the city of Seoul in the years ahead.

Development and Application of a Deep Learning–Based Real-Time Urban Flood Monitoring System

목록