Practical Data Visualization and Virtual Reality Virtual ...webstaff.itn.liu.se/~jimjo/courses/TNM093-2013/lectures/TNM093-le… · VR Interaction and Navigation Karljohan Lundin

Practical Data Visualization and Virtual Reality

Virtual RealityVR Interaction and Navigation

Karljohan Lundin Palmerius

Synopsis

● Tracking● Navigation● Interaction● Mixing real and virtual (AR)

Tracking

● Head position (head tracking)– Parallax effects

– Correct 3D sound

– Natural Navigation● Moving around objects● HMD view

● Interaction– 3D mouse

– wand devices

Degrees of Freedom

● Tracked dimensionality– 3 DoF – position x, y, z

– 3 DoF – orientation (”yaw, pitch, roll”)

– 6 DoF – position and orientation

– More variations● 2 DoF, 5 DoF, 12 DoF, etc

● From lower to higher DoF– Two 3 DoF becomes 5 DoF

– Three 3 DoF becomes 6 DoF

Dead Reckoning

● Absolute tracking– grounded to the room

– calibrated to world coordinates

● Dead reckoning tracking– accelerometers and gyros

– know the acceleration and change of orientation

– integrate to get velocity and again to get position

– accumulates noise and integration errors

● Hybrid Tracking– combining to get the best of two worlds

Intersense● Hybrid tracking

– Ultrasound● slow but absolute● speakers and receivers● distances by time-of-flight● position by triangulation

– Accelerometers● fast but dead reckoning● accumulates errors over time

– integration error ( a → v → x )– noise and inaccuracy

Electromagnetic Trackers

● Magnetic field generator– Fast switch between magnetic fields

● Distributed detectors– Detect local shape of field

and compare the generated fields

Camera-based Tracking● Colour or IR cameras● Detect features

– Reflector balls – two cameras: 2D line into 3D position

– Face – search for skin colour and distribution

– Markers – 6 DoF from perspective maths

Characteristics

● Scalable– Cheap or expensive cameras

– More cameras for better tracking

– More or less complicated algorithms

● Issues– Frequency and latency (CPU expensive)

– Precision

Depth Camera Tracking

● Image-based tracking– like markerless tracking

– depth data makes things easier

– several approaches● active w/ structured light● active w/ time-of-flight● passive w/ image disparity

Depth Camera Characteristics

● Accessible– Cheap

– Available algorithms and implementations

– Need little tweaking

● Issues– structured light cameras are slow

– often low resolution because of CPU strain

– active approach is sensitive to bright light (IR)

– active approach cameras interfere each other

Consumer Grade Tracking

● Wii Mote● Playstation Move● PrimeSense

– depth camera (e.g. Kinect)

● LEAP motion● Myo

Mouse

● Magellan SpaceMouse– Control velocity not position

● Interaction– Mouse pointer

– Control objectr transform

– Navigate

● 6 DoF velocity control is hard to use– Reduce DoF or use major axis

Wand/Stylus

● Extension of 2D mouse– Real 3D position control

– Co-located with VR world

● 3–6 DoF in usable form– Hybrid tracking

– Mechanical tracking

● Interaction– Pointing device

– Touch, pick, grab

– Buttons

Implicit Navigation

● Immersive display– Should give us ‘real world’ interaction

– Walk to the object, turn around, etc

● Dependent on display and tracking technology– Caves and HMDs vs workbench and workstation

– Available space

– Locomotion interface Tracked volume

Screen

Explicit Navigation

● Select where to fly and how● Typical metaphors

– Fly/walk direction

– Goal driven navigation

– Object driven navigation

Fly/walk Direction

● Gaze direction– Natural feeling

– Can't watch scene go by

– Rare in VR, common in FPS

● Pointing mode– Point to desired position

– 5 DoF device tracking required

● Crosshair mode– Eye/hand line defines direction

Hand Controlled Speed

● Distance– Hand motion relative to initial point

– Dead zone is nice

● Used with crosshair control– Intuitive

– Natural mapping

– Limited range

Goal Driven Navigation

● Fly-to– Click on pre-defined ”book marks”

– Magic telescope

● Automatic navigation– Pre-selected ”good” path

● Virtual map– Select position you want to be at

– Type of widget

Object Driven Navigation

● Lifts, stairways, teleporters– Objects that transports you

– Instant or soft motion

● Exotic variations– Attractors, repellors

● Need to know where people want to go– Virtual architecture

Situation of Interaction

Tracked volume

Screen

Selection Calculations● Image space

– Don't use GL-based ”pick”● Not one image to work in● Generally not 2D

– Not always eye oriented

● VR world coordinates– Linear algebra

● Closest● Intersection

– Explicit handling of coordinates● Check your frame of reference!

Frames of Reference

● VR display system– graphics origin frame

– eye position

● Navigational system– graphics origin offset

– move objects relative the display

– move the display relative objects

T

TNAV

Scene

UI

Selection

● Close up– Touch metaphor

● Walk to object, reach out and touch● Impossible if object is behind the screen

– Encircle with line

– Select with pointer/sphere

● Far away– need extended reach

Extending Your Reach

● Pointer metaphor– Point at objects to select

● 5 DoF tracker required

– Issues in densly populated scenes● Select the closest● Occlusion

Extending Your Reach

● Push workspace– Move workspace when device is at its border

● Visual indication● Haptic indication

– Good for small workspace

– No co-located interaction

Extending Your Reach

● Mouse pointer on a stick– Extendable stick

– The point is the active pointer

– Control distance● Automatic, joystick, gestures

Extending Your Reach

● Typical Issues– No co-location

● Less intuitive● Less natural and effective● Possibly less precision

– Densly populated space● Find the right object● Hit the right object (occlusion)

– Transformed workspace● Never, ever rotate the workspace● Help user to track changes

Reality/Virtuality Continuum

● Mix between real and virtual world– ”Milgram's Continuum”

Real Environment Virtual EnvironmentMixed Reality

Augmented Reality Augmented Virtuality

Reality/Virtuality Continuum

● Mixed Reality– Anything with real and virtual components

● Augmented Virtuality– Virtual Reality augmented with live video feed

● Augmented Reality (AR)– Mostly real environment

– Augmented with computer graphics and VR

Augmented Reality

● Three key elements (Azuma)– combination of real and virtual reality

– real-time updates at interactive speed

– co-registration of real and virtual objects

Applications

● Annotation– Guided tours

– Sports

– Manufacturing and maintenance● Boeing - length and bundeling of cables● BMW - maintenance of engine

Applications

● Augmented Vision– ”X-ray vision” in medicine

– Building plans for smoke divers

UNC Chapel Hill

Applications

● Modifying Reality– Architectural or design modifications

– Virtual hairdressing / makeup

Applications

● Games– Augmented conventional games

– Monsters in reality

– Mark out real people as friends & foes

Special considerations

● Graphics– Need of fast updates

– Mixing of real and virtual visuals

● Tracking– High demands on precision, accuracy, latency

– Need of tracking of the world

Virtual/Real World Interaction

● Lighting, lamps, shadows, occlusion● Collision, co-location, manipulation