Development of Real-Time Gesture Recognition System Using Visual Interaction

摘要: The aim of this paper is to present a methodology for hand detection, propose finger detection method, and finally apply them posture recognition. detected can be used implement the non-contact mouse. This technology control home devices such as curtain television. Skin color segment region from background, counter extracted segmented hand. analysis gives us location fingertip in Fingertip tracking performed assuming constant velocity model using pixel labeling approach. From process, we extract several features that are fed finite state classifier identifies configuration. classified into many gesture classes or different movement directions. method skin segmentation assumes background does not contain any colored object beside hands. We have an extensive experiment achieved very encouraging result. Ultimately, suggests empirical application verify adequacy validity proposed systems. Accordingly, satisfaction quality services will improved