The Innovation Development of Early Flash Flood Warning System Based on Digital Image Processing through Android Smartphone

Abstract

Abstract Flooding is a phenomenon that almost always occurs every year in various regions throughout Indonesia with a variety of cycle variations. In the rainy season, high rainfall and narrowed watersheds and silting up are no longer able to accommodate the flow of water, causing flash floods. Flood Warning Systems operate with traditional local wisdom that is highly dependent on the availability and responsiveness of the personnel assigned to oversee the river. Therefore, the development innovation of the existing system is expected to be able to provide information and data regularly and be an early warning to flash floods in order to minimize the impact. This research was conducted to design, create and implement an image-based flash flood early warning system prototype based on an Android cellular phone so that the public is expected to be able to access information on possible flash floods through their respective cell phones. Image processing is the basis of this development innovation through the Background Substraction method which is carried out to automatically detect High Levels of Water (TMA) on the control board (Peil) via an IP Camera which is an input parameter for the system and is processed into information for interested parties. The results of analysis of the detection of water level (TMA) on the peil board carried out on the server computer, become input parameters for determining the status of the dam water level. Flood early warning application that is installed on Android smartphone devices receives data from the server computer via the internet network.