Application of Geospatial and ICT Technologies for Landslide Disaster Risk Reduction in Rwanda

Application of Geospatial and ICT Technologies for Landslide Disaster Risk Reduction in Rwanda

English

2590-0617

This study provides a new insight into landslide advancement based on the state-of-art mapping technology, time-series imagery, and disaster informatics in Rwanda, where 40 % of its lands susceptible to landslides. This is the first nationally supported and locally evidenced use-case of the iMaster/DocuCam, a scalable system for landslide monitoring and detection in the African continent, originally developed for advanced industrial automation. The system continuously collects high-resolution images and weather data to monitor landslides and alert relevant authorities. We used “Quadtree” Computer Vision algorithms to automatically detect landslides by pre-processing images and measuring brownish pixels. Using statistical modelling, comparison with previous images is made and deviation in the number of brown pixels is determined. If deviation exceeds the threshold, a landslide is detected, and a warning is sent out. The results are also stored in the Open-Source SQL database and accessible via web-based Graphical User Interface. We also explored user requirements, needs and demands of stakeholders at national and local level for developing the people-centered early warning system in Rwanda, as promoted by the Sendai Framework. This study manifests the commitments of the Rwandan government to accelerate the localized disaster risk reduction action for societal resilience in the tropics.

ART 636

Progress in Disaster Science, Vol.28, No. 100466, December 2025

eng

ANA-131401