A Training Gesture-Based-Scroll Visual Artificial Intelligence And Measuring Its Effectiveness Using Hidden-Markov Modeling Methods

Abstract

In this article I discuss the method of hand gesture recognition as a visual motion detection based on artificial intelligence by training three main movements namely, scrolling up, scrolling down and stopping based on capturing the front camera image capture speed of 3 fps and measuring its efficiency against the control movements that performed using Hidden-Markov Modeling (HMM) with each catch object scroll up 3 fps / 15 frames scroll down 3 fps / 15 frames and stop 3 fps / 9 frames, the result is that the most effective hand gesture object training movement is stopped gesture with 3 fps / 9 frames because the object's movement is able to be recognized by the system only in the 3rd-second image capture frame.

Keywords

• gesture-based-scroll
• visual artificial intelligence
• hidden-markov modeling