A camera transmits a captured image of a surveillance area to a server. A microphone receives a sound of the surveillance area and transmits the sound to the server. The server detects a warning sound starting based on a harming behavior on a vehicle in a surveillance area and arriving from the vehicle using the sound of the surveillance area, determines an occurrence position of the harming behavior based on the detection of the warning sound, acquires a captured image of the occurrence position, determines the vehicle on which the harming behavior is executed using the captured image of the occurrence position, acquires information regarding the vehicle, and records the information regarding the vehicle and the captured image of the occurrence position in association with each other in an accumulator.