IVR Chapter 7 Motion tracking

1. Measures as position, orientation, angle, acceleration, distances

2. Dimensions // 3D for VR

3. Number of tracked segments

4. Data handling automatically or manually and in real time or not

5. Size of interaction volume (tradoff)

6. Ergonomics (intrusive or not)

1. Spatial accuracy (Grasping vs walking) // Hit around middle

2. Spatial precision // Hit exact anywhere

1. Temporal accuracy (Response time, latency, sample rate)

2. Temporal precision (Jitter)

1. Mechanical with the ultimate displayed

2. Electromagnetic

3. Instrumented gloves

4. Ultra sound

5. Intersense

6. Light

7. Satellite with global positioning system (GPS)

1. Polhemus

2. Ascension

3. Used in military

1. Good position and orientation

2. Sensitive to ferromagnetic materials in the environment

3. Accuracy decreases with growing distance

4. Large sensors (1-2cm)

5. Cabling

1. Measures angles between finger segments

2. Difficult to measure thumb rotation for precision grip

1. Only makes sense with other trackers

2. Different qualities with flexion vs extension and adduction vs abduction

3. Error propagation

4. Discomfort

5. Calibration process necessary

1. Inertial and acoustic

2. Combination of accelerometers and gyroscopes

1. Time of flight sensors

2. Structured light (e.g. Mircosoft Kinect)

3. Opto-electronical (e.g. A.R.T.)

4. HTC Vive with lighthouse

5. Oculus rift

1. Measure distance between camera and objects

2. 200x200pixels with 100Hz resolution

3. 1cm accuracy

1. Pre-defined matrix of points by IR laser projector

2. Deformed point matrix captured by IR camera

3. Texture mapping by VGA camera

1. Use triangulation for 3D conversion

2. 640x480pixels with 30Hz resolution

1. 2 or more IR camers track 2D position of markers

2. Calculate 3D position via direct linear transformation (DLT)

3. Passive with IR light reflecting markers

4. Active with IR light emitting markers with cabling

1. Selspot experiment of grasping real and virtual object

2. Good accuracy and latency sample rate

3. Occlusion possible

1. Two basis stations with IR LED lasers

2. IR light sensors on the HMD front

3. Time differences yield position and orientation

Also possible with one basis station with limited view angle

1. Combination of gyrometer, accelerometer and magnetometer for R

2. IR LEDs attached to HMD and IR camera for p

Get x_i with direct measure y_i = Rx_i + p and rotation R and translation p

1. x_m = 1/n * Sum_i=1ⁿ x_m and y_m = 1/n * Sum_i=1ⁿ y_m

2. X = [(x₁-x_m),…,(x_n-x_m)] and Y = [(y₁-y_m),…,(y_n-y_m)]

3. C = Y*X^T the covariance matrix

4. C = U*W*V^T with sigular value decomposition

5. R = U*(1,0,0;0,1,0;0,0,det(UV^T))*V^T

6. p = y_m - Rx_m