Abstract
Although there are many researches regarding strip roughness effects on flow resistances, researches under shallow flow conditions have been rarely conducted. In this study, we conducted shallow water experiments by varying roughness spacing, bed slope and the Reynolds number, and categorized water surface patterns into 4 patterns, that is, Flat surface, Nomal Jump, Periodic hydraulic Jump, and 3-Dimensional standing wave. In addition, we successfully showed region segmentation figures for each water surface patterns by using SVM(Support Vector Machine) algorithm. This approach enables us to estimate water surface patterns approximately.
