Cosc 6326/Psych6750X

Enabling Technology for Advanced Displays

• Virtual reality and other advanced interactive displays– simulate and maintain a model of the world to

be created or augmented– present or display the world to the user

(displays and effectors)– sense the actions of the user and environmental

state to enable the simulation to react (sensors)

• A typical VR system has – sensors to collect information about the actions

of the user – a processor to collect this information, model

the virtual world and generate the output for the display devices.

– displays and other sensory stimulators generate the sensory input provided to the user.

• The sensation-perception-action cycle of the user is an integral part of VR system.

• Normally when one acts in the world feedback from the senses confirms the expected result

• Current VR systems have serious limitations that limit the ability to create high fidelity realistic virtual worlds.

• In a sense VR closes the users sensorimotor loop– User acts in the world. – Simulation detects the action using sensors– Feedback provided by the simulation via the

displays

1. Simulation and Image/Display Generation

• Hardware– Need to provide real time update to the user– Processor speeds and technology have

improved exponentially although modelled VR worlds are still limited

– Recent trend move from ‘big-iron’ to clusters of PCs

• Software:– input, simulation, rendering– often done in parallel loops (more

parallelization possible)– input loop handles interfacing with sensors to

get current state

– simulation loop: • for each time interval simulate behaviour of objects

in virtual environment• physical behaviour, reaction to user actions, higher

level behaviour (intelligent entities, avatars …), collision detection …

• real time – feedback to user must be timely (e.g. 60 Hz)

• distributed, multiprocessor pipelines …

– Rendering loop:• generate displays to present: graphics, haptics, audio• modern raster graphics has a number of stages to convert world

model to raster image– transformation, projection– lighting, shading– texture mapping – rasterisation– anti-aliasing– visibility, clipping, culling …

• recently, substantial hardware support on fast, low cost graphics cards – ‘graphics pipeline’

2. Displays and Effectors

Low end HMDs

• Targetted for personal entertainment (games, dvd, …)

• Sony Glasstron, Olympus Eyetrek• currently NTSC, PAL, VGA resolution. HDTV?

VR HMDs

• Sutherland’s HMD was boom supported. Often need free head motion.

• Characterizing HMDs– Configuration: projection versus direct viewing– Optics: simple magnifier vs. compound

microscope– Display image source: CRT, DLP, LCD …– Opaque or see-through

VR HMD

Projection type • head mounted optics• external electronics &

projection display • CAE FOHMD

– images generated by high-resolution data projectors

– coherent fibre optic bundle and optics direct image to eyes

• Direct viewing: many modern displays have head mounted miniature displays– CRT: e.g. N-Vision, Kaiser (KEO)– LCD: e.g. Virtual Research, KEO– laser retinal scanning– DMDs – FEDs …

Some HMDs

HMD Optics• Simple magnifer

– single magnifying lens, short optical path– no exit pupil formed– simple, inexpensive

• Compound optics– several lens: eyepiece, objective– exit pupil formed; must align with eye’s entrance pupil– more complicated, longer optical path, permits focusing

See-through HMD capability

• Non-see-through– No need for optical combiner– Eye sees only the virtual image – Pure virtual reality application

See-through HMD capability

• Optical see-through– images of the real and virtual worlds optically

superimposed – need optical combiner (transmission ratio?)– useful for AR, wearables; similar technology

for heads-up displays– distortions and time-lags a problem– direct view of real world

See-through HMD capability

• Video See-through– non-see-through HMD plus ‘scene’ cameras– the virtual world is superimposed on a video

image of the real world– electronic (not optical) combiner– can match time delays and distortions– system has access to user’s view– low resolution image of the real world

• Figures of Merit/Design factors– field of view– resolution (tradeoff with FOV)– luminance, contrast– colour– monocular, biocular, binocular – exit pupil size, eye relief, adjustments (inter-pupillary

distance, focus)– distortion

Projection-based displays

• Walls– large screen interactive displays– suggested for collaborative design– curved screen, flat, wrap around, dome

• e.g. Elumens Vision Dome

• Desks– ImmersaDesk (University of Illinois EVL), …

CAVE/CAVE-like• University of Illinois EVL, Fakespace,

Trimension (ReaCToR), Mechdyne (SSVR )

from fakespace

Cave

• reconfigurable CAVE - RAVE

Large format immersive displays

• Large format film, domes, planetariums, ride simulators– SEOS, Trimension, Spitz, Disney Quest, IMAX– immersive but often not very interactive (large

groups)– used in simulators, $$$ for VR– Mechdyne V-dome has been used for VR

V-Dome

• projection technology issues– projectors

• cathode ray tube (CRT)• digital light processing (DLP)• DILA• liquid crystal display (LCD)• Laser

– screens• material: glass, fabric, plastic, fog!• reflectivity, gain, polarisation• inter-reflection (black screens)

– structure• single vs multiple• tiling, blending• colour and luminance matching/uniformity

– support for stereopsis

• Audio displays:– stereophonic, surround sound– spatial sound displays– sound modeling and synthesis

• haptics, tactile displays …• more on these later …

Sensors

• Sensor technology is currently particularly rudimentary.

• Position of a limited number of joints or limbs is normally sensed such as the position of the head and hand.

• Buttons and joysticks etc can also provide input.

• Sense only a limited range of the possible motions and have limited resolution.

• Lag is a major problem with some sensors

Tracking Technology

• To generate the displays, need to know users position and orientation

• Need to track user’s head (hand, body …) in real time in order to respond to head (hand, body …) motion in real time

• Current tracking does not measure degrees of freedom possible in human motion

• magnetic– pulsed DC, AC– earths magnetic field

• ultrasound• optical• GPS (outside)• mechanical• gyroscopes, accelerometers

www.3rdtech.com

3D input devices

• a number of 6 degree of freedom input have been proposed for 3D interaction

• spaceball, 3D mice, hand/stylus tracking

• isometric versus isotonic– maps to rate versus position

control

Gloves/Motion capture

• one of the early VR input devices was the Dataglove

• typically many degrees of freedom

• additional tracking for position

• animation/gesture recognition

Gypsy

ImmersionCybergrasp

Other input technology

• speech recognition• eye gaze tracking• gesture recognition• biopotentials