In this research, a Flood Warning System is designed based on Water Level Data image obtained by a simple web camera. The system processes the output data from the water level image into tables and then connect to the flood map database to select the corresponding map to be displayed as a flood prediction image with a GIS program. The system processes the data in real time to provide public users with an accurate flood area prediction. By using this system, user can expect the area which will be affected when the water level rises to allow peoples that live near the affected area to evacuate or at least to prepare for the upcoming flood. This kind of early warning system is expected not only will save the lives of people who live near the affected area but also save their valuables from the flood disaster. The flood area is segmented from LiDAR data at 1 m intervals and the water level is recorded at 10 cm intervals. The resulting area obtained from elevation data is considered as the boundary of the maximum extent of the flood. Each corresponding map layer is linked to specific water level recorded by the camera.