Cambridge-MIT View Sequential Display

Oliver Cossairt, Christian Moller, Steve Benton, Adrian Travis

Abstract

View-sequential 3D displays are used to create images whose appearance changes with viewing position so that scenes can be displayed with motion parallax, which is an important depth cue that greatly improves the apparent realism of displayed images. View-sequential displays utilize a time-multiplexing principle to display images at different viewing directions sequentially in time. The concept of view-sequential displays was first demonstrated at the Cambridge University Photonics and Sensors Lab. In 2002, Cambridge University and MIT collaborated to build a view-sequential 3D display using a Texas Instrument DMD™ in conjunction with a fast switching ferro-electric shutter. The display is capable of producing 16 views at 15-bit color with a refresh rate of 50Hz. In 2004, Christian Moller founded the company Setred to commercialize view-sequential 3D displays for Medical Imaging and Oil and Gas Visualization.

Publication



Display Optics



View-Sequential display optics consist of a pixel generating device, an objective (projection) lens, an active shutter, and a field lens. For the Cambridge-MIT display, a 3-chip DMD projector is used to generate images and an custom designed FLCD for the shutter. The FLCD and DMD are closely synchronized so that a one 15-bit color image is displayed while a single shutter element is open on the FLCD. A sequence of 16 images/shutter pairs are updated in 1/50th of a second.

Images

These are images taken from a single viewpoint of 3D objects displayed on the the view-sequential display.


A 3D image of a CG aeroplane
A 3D image of a CG butterfly

Videos

These are videos showing motion parallax for 3D objects displayed on the view-sequential display.


A 3D image of a CG Volkswagen Beetle
A 3D image of a CG Curch