Virtual Volume Control-By-Hand-Indication-

Hand gestures are a hot topic in the vision community, with the sign language recognition and human-computer interface applications. One of the first tracking systems focused on articulated hand movements. Their technology, which retrieved point and line data from grayscale pictures, allowed them to monitor a 27-degree-of freedom hand at 10Hz. However, it has trouble tracking in the presence of occlusions and complex backdrops, and it requires a human initialization step before it can start monitoring. The majority of hand tracking research has thus far concentrated on 2D interfaces from an interaction standpoint [18] and used low-cost web cams to follow a finger around a planar region to handle a standard graphical interface without requiring a mouse or keyboard. Fingertip identification was achieved by fitting a conic to rounded features, and local tip tracking was accomplished with Kalman filtering. Similarly, infrared cameras were utilized to divide skin regions from background pixels to track two hands on a 2D tabletop display for interaction. After that, their system used a template matching approach to recognize a small set of motions that might be translated as interface commands.