User-Centered Artificial Intelligence for High Spatial Urban Flood Mapping
Loading...
Date
2023-05-11
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The number and intensity of flood events have been on the rise in many regions of the world. In some parts of the U.S., residential properties, transportation networks, and major infrastructure are at risk of failure caused by floods. The vulnerability to flooding, particularly in coastal areas and among marginalized populations is expected to increase as the climate continues to change, thus necessitating more effective flood management practices that consider various data modalities and innovative approaches to monitor and communicate flood risk. With the goal of improving the quality of decisions made during flood events to evacuate people or move goods and services while avoiding flooded areas, this study first conducted a community needs assessment that highlighted the need for real-time data on floodwater conditions, risk-informed evacuation plans, and safe and shortest transit routes. Research points to the importance of reliable information about the movement of floodwater as a critical decision-making parameter in flood evacuation and emergency response. Existing flood mapping systems, however, rely on sparsely installed flood gauges that lack sufficient spatial granularity for precise characterization of flood risk in populated urban areas. To address these needs, this study utilizes crowdsourced photos of submerged stop signs to generate street-level flood maps with high spatial resolution (compared to flood gauge data), and develop risk-informed evacuation plans during floods by comparing sign’s pole length in a paired photo (before and after a flood). To enable crowdsourced data collection, a web application, called Blupix, is designed and launched. Images collected by the Blupix app are analyzed using two developed computer vision models (Computer Vision Model I and Computer Vision Model II) and validated. Next, two route optimization algorithms are implemented and compared based on the generated inundation map to determine the shortest flood-free evacuation route. Next, a mobile application (called Blupix Mobile) is developed with embedded computing functionality for on-demand floodwater depth estimation from geocoded photos of submerged stop signs which provide ordinary people and first response teams with reliable, ad-hoc flood water depth information. Finally, a user study is conducted to assess human’s perception of flood risk using immersive virtual reality (VR).
Description
Keywords
Artificial Intelligence, Computer Vision, Image Processing, Convolutional Neural Networks, Decision Support System, Crowdsourcing, Route Optimization, Disaster Management, Floods, Flood Mapping, Flood Risk Perception, Mobile Application, Web Application, Immersive Virtual Reality, User Study, Community Engagement