Recognition of hand gestures on the video stream based on a statistical algorithm with pre-treatment

Abstract

The goal of this work is human hand detection and gesture recognition. This is a tremendously difficult task as hands can be very varied in shape and viewpoint, they can de open or closed, they can have different finger articulations. It is proposed a combined method of hand gesture recognition based on a statistical image processing algorithm. As a pre-processing algorithm it was applied Lucas-Kanade method and background subtraction algorithm. Object recognition was performed with using of Haar classifiers. The possibility of using gestures for remote control of various devices with different hands position from the camera location was shown.