Dal Alert!

Receive alerts from Dalhousie by text message.


Honours Thesis Presentation - Unassisted Holographic Projection Using AR, Face Tracking and Motion Detection Technologies

Who: Rui Xu


Title: Unassisted Holographic Projection Using AR, Face Tracking and Motion Detection Technologies

Supervisor: Dr. Stephen Brooks

Reader: Dr. Qigang Gao


The Unassisted Holographic Projection display offers the means of visualizing 3D contents without the use of special glasses, head mounted displays or any handheld devices. The Unassisted Holographic Projection System projects a 3D space that the user can interact with. To give a realistic stereogram illusion to the user, the system automatically corrects the viewpoint by tracking the user’s position and analyzing their eye movement using an off-the-shelf web camera. Hand interactions are also facilitated by integrating this with a standard Microsoft Kinect Motion sensor.

The motivation of this project is to solve problems that traditional AR, VR and autostereoscopic displays have, such as virtual reality sickness, lack of mobility and high levels of dependency on handheld or wearable devices. Compared to autostereoscopic displays, the Unassisted Holographic Projection System gains advantages in degrees of freedom and mobility.  This is because current commercially available autostereoscopic displays are based on lenticular sheets or parallax barriers and have numerous limitations; in particular, the user cannot move their hand around in the “3D space” as it’s displayed on a 2D surface. This thesis will provide a more complete understanding of this new display technique, with an introduction of the background and history, and a comparison between existing holographic projection methods and AR/VR systems. Furthermore, the technologies, algorithms that form the basis of my system and a subsequent performance analysis are also presented in this thesis.  Lastly, opportunities for further studies and improvements are stated.



Room 311, Goldberg Computer Science Building