A system for inputting operation system (OS) commands to a data processing device. The system comprises a video camera that captures images of a viewing space. A processor detects a predetermined object in the images using an object recognition algorithm not involving background information in an image. One or more image analysis parameters of the object are extracted from the images and one or more motion detection tests are applied. Each motion detection test has an associated OS command, and when a test succeeds, the OS command associated with the test is executed. By not relying on background information in an image, the system of the invention may be used in devices that are moved in use, such as a palm plot, personal digital assistant (PDA), a mobile telephone, a digital camera, and a mobile game machine.