In recent years, climate change has led to a global increase in natural disasters, resulting in significant loss of life and economic damage each year. Notably, more than 80% of flood-related disasters occur in river and small stream basins; however, there is a severe shortage of personnel and equipment to properly manage these areas. In response, the Ministry of the Interior and Safety of Korea has been promoting the establishment of a "Smart Measurement and Management System" since 2023 to prevent disasters in small streams caused by climate change. Based on this initiative, technologies capable of forecasting and measuring on-site conditions using Artificial Intelligence (AI) are currently under development. This paper introduces the “Development of a Video-Based Flow Measurement System Using Deep Learning Models and an Automatic Velocity Vector Adjustment Algorithm” which is currently being carried out by our institution. The outcomes of this research are expected to support disaster response even in unpredictable situations, thereby reducing social and economic losses. Furthermore, if applied to the management of small-scale rivers and streams, the technology could significantly contribute to flood prevention and mitigation efforts by local governments.