This solution is a framework in which some skin patterns are selected by the user in order to build a recognition module to be applied in video scenes, e.g., captured from a simple webcam device. Once these patterns are defined on the scene, tracking algorithms are used to determine the movement of the regions of interest (such as face, arms and hands). This kind of skin recognition can be used as a human-computer interaction method and therefore enabling the user to control a computational environment in a more natural manner.