In the following device, at the time of pan/tilt movement of a pan head, images #1 to #3 for three frames shot by a camera are sequentially taken in into an image memory of an image processing device. And, processing of an image processing processor as well as a CPU compares the image #1 with the image #2 and the image #1 with the image #3 respectively so as to extract an image of an moving object so as to detect moving velocity of the moving object based on the image of this extracted moving object. When the moving velocity of the moving object is detected, a command signal to cause the pan head to pan/tilt at the moving of that moving object is given to a pan head controller by the CPU. This moves the pan head in the pan/tilt direction at the moving velocity of the moving object. Thus, detection of the moving velocity of the moving object from the shooting image of the camera and control of the moving velocity of the pan head based on the detected moving velocity of the moving object enables following on the moving object without fail.