With the growing digitalization of our society and the increasing development of Industry 4.0, the development of safe human-computer interfaces is of prime importance. For this purpose, we present a new solution for avoiding any physical contact between the human and the machine’s display and/or its keyboard/pad. Hence, we developed an innovative image processing algorithm which runs on the live video stream of the remote interaction between the human and the machine such as a computer or an intelligent agent; the video stream being captured by an off-the-shelf camera which is embedded in the main machine. The resulting computer-vision system performs visual object detection and tracking of a marker the user holds in one hand and could wave it from a distance in order to remotely interact with the interface of the machine. This remote interaction system has been successfully tested in real-world operational conditions, and the performance of this image processing application has been evaluated in real time.