2021 Volume 77 Issue 2 Pages I_325-I_330
In the Hojo River, which flows through central Tottori Prefecture, the flow rate during rainfall increases. Therefore, inundation damage occurs frequently. The river manager judges the river condition by checking the CCTV camera image installed in order to carry out flood control activities promptly when the water overflows. However, there is a shortage of personnel, making continuous monitoring difficult. Therefore, there is a problem that it is not possible to grasp when and where the flood occurred.
In this study, we developed a model that enables automation of overflow detection using deep learning of artificial intelligence with advanced image analysis ability.In addition, we confirmed the usefulness of the model through field demonstration experiments aimed at improving overflow monitoring. The model developed in this study may be applicable to CCTV cameras installed for river monitoring.
PROCEEDINGS OF HYDRAULIC ENGINEERING
