Image processing for flood detection and prediction

• Grant number: 21/10921-2
• Support type: Scholarships in Brazil - Post-Doctorate
• Effective date (Start): December 01, 2021
• Effective date (End): November 30, 2022
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computer Systems
• Principal researcher: Jó Ueyama
• Grantee: Caetano Mazzoni Ranieri
• Home Institution: Instituto de Ciências Matemáticas e de Computação (ICMC). Universidade de São Paulo (USP). São Carlos , SP, Brazil
• Associated research grant: 13/07375-0 - CeMEAI - Center for Mathematical Sciences Applied to Industry, AP.CEPID

Abstract

Floods have been problematic in Brazil often causing material and life losses. Several cities like São Paulo, São Carlos, Osasco and Mogi das Cruzes have suffered from the floods. The group responsible for this research carried out relevant work in the area of flood detection using technologies such as the Internet of Things. Flood detection sensors were installed in the city of São Carlos in partnership with the Town Hall of São Carlos-SP, Brazil. This project aims to detect and predict flooding using images and thus automate the process of flood identification without human intervention. Such an approach only uses cameras without the need for the river height sensor that remains submerged in the urban streams. River height sensors are usually susceptible to failure as they are continually in contact with river water. In addition, Civil Defense bodies often require flooded river images and therefore we believe that the use of image processing for flood detection is timely as one single sensor (i.e. a camera) is needed to detect floods and provide images to Civil Defense bodies. Different models for flood prediction based on river-side cameras will be designed, and contextualized analyses considering different types of modalities will be performed. (AU)