Streaming and Interactive Exploration of Dense Points Clouds in Immersive Virtual Reality

Master Thesis

About This Topic

In recent years, 3D mesh generation using depth sensors like Microsoft Kinect got very popular and a number a algorithms exist for real time depth data fusion. However, exploration of this mesh data is mostly done using standard desktop-based viewing tools missing intuitive means of exploring the mesh by travel using natural walking.

In this thesis, a research prototype is developed for interactive exploration of a dense 3D mesh using an immersive virtual reality setup. The depth data is collected using a notebook and a connected MS Kinect and streamed to a workstation that transforms the 3D point cloud into a mesh. The 3D mesh can be interactively explored while being constantly updated over the streaming component using a omni-directional treadmill (Cyberith Virtualizer) for natural walking and a HMD (Oculus Rift) for egocentric scene view.

The topic is ideal to gain a deeper understanding in immersive virtual reality, VR hardware and 3D human computer interaction.

Programming: C/C++, C#

SDKs/Platforms: Point Cloud Library, Unity3D, Unreal, Virtualizer, Oculus Rift