Parking assist devices, methods, and programs determine a current position of a vehicle and acquire image data from a camera provided on the vehicle. The devices, methods, and programs extract a predetermined region of the acquired image data, the predetermined region being a portion of the image that will be in a blind spot of the camera if the vehicle travels a first predetermined distance from the current position. The devices, methods, and programs correlate the extracted image data with the current position of the vehicle and cause the correlated extracted image data to be stored in a memory. The devices, methods, and programs read out past correlated extracted image data, the past correlated extracted image data having been acquired when the vehicle was in a previous position, the previous position being a second predetermined distance from the current position. The devices, methods, and programs output a parking assist image to a display, the parking assist image displaying both the past correlated extracted image data, and the image data acquired at the current position of the vehicle.