In this study, a closed-circuit television (CCTV) system, installed for surveillance purposes, is utilized to measure the flow rate during a flood. The procedure to determine both the angle and scale-factor of the camera is described. Then, image analysis techniques, namely the direct visual measurement method, Large-Scale PIV (LSPIV) and Space-Time Image Velocimetry (STIV), are applied to the video images recorded by the CCTV camera. The results of these methods and the conventional float measurement are compared. In addition, the accuracy of the respective methods is discussed. A set of low-quality video images of a flood during a thunderstorm that occurred under the dark ambient conditions (midnight) is analyzed using three image-based methods. The transition of the flow rate during the event is successfully estimated.