Latest revision as of 12:26, 1 July 2024

Abstract

This paper presents a development of land use classification model based on semantic segmentation using aerial photographs and its contribution to the efficiency of 2D tsunami inundation simulations. The proposed method uses Artificial Intelligence(AI)-based image classification to generate a roughness coefficient mesh, which is then applied to a 2-D tsunami run-up simulation using the finite element method on real geometry. Numerical results are compared to evaluate the improvement in simulation efficiency, and the potential benefits of the proposed method are discussed by analyzing the differences in simulation results.

Full Paper