An optimised 3D view of Kinect live stream

Abstract

This paper describes a method to generate anaglyph images based on a single colourful captured frame and the depth values provided by Microsoft Kinect. The final result can be viewed using 3D Red/Cyan anaglyph glasses. This technique will reduce the information process on the digital contents to generate 3D frames from a single RGB source. The proposed technique provides a 3D view using 2D images where no 3D modelling is needed. It also has the advantage over the recent approaches were the 3D images are generated using two RGB input sources. The contribution in this system is the ability to work in real-time as a 3D video stream. It also helps the users with a better sense of immersion and improves the naturalness of the interaction in comparison to the 2D interaction.