Top Banner
Development of an Eye Development of an Eye Tracker Tracker By Jason Kajon Barrett By Jason Kajon Barrett of the Center for Imaging of the Center for Imaging Science Science at the Rochester at the Rochester Institute of Technology Institute of Technology Advisor: Jeff Pelz Advisor: Jeff Pelz
26

Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

Dec 22, 2015

Download

Documents

Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

Development of an Eye TrackerDevelopment of an Eye Tracker

By Jason Kajon BarrettBy Jason Kajon Barrett

of the Center for Imaging Scienceof the Center for Imaging Science

at the Rochester Institute of at the Rochester Institute of TechnologyTechnology

Advisor: Jeff PelzAdvisor: Jeff Pelz

Page 2: Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

What is an Eye Tracker?What is an Eye Tracker? A device used to detect eye motion.A device used to detect eye motion.

What are they for ?What are they for ?

• Access for the physically handicapped, Access for the physically handicapped,

• target variable resolution,target variable resolution,

• realistic virtual reality environments,realistic virtual reality environments,

• and increased bandwidth of human-computer and increased bandwidth of human-computer interaction. Shannon’s Rule: C = BW log interaction. Shannon’s Rule: C = BW log22(1 + S/N)(1 + S/N)

Page 3: Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

Types of TrackersTypes of Trackers: :

• mechanical- corneal covermechanical- corneal cover

• magnetic- eyeball polaritymagnetic- eyeball polarity

• electric- EMG & EOGelectric- EMG & EOG

• optical- IR illuminationoptical- IR illumination

Page 4: Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

What is the Research Design?What is the Research Design?

A device using four IR photodetectors and a A device using four IR photodetectors and a microcontroller to detect the onset of rapid eye microcontroller to detect the onset of rapid eye motion. A PicBasic program to run this device.motion. A PicBasic program to run this device.

Page 5: Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

Why a microcontroller ?Why a microcontroller ?

• Simple data manipulation, Simple data manipulation,

• lookup table,lookup table,

• single upgradable chip,single upgradable chip,

• cost for device,cost for device,

• and ease burden on CPU.and ease burden on CPU.

Page 6: Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

Targets for the ResearchTargets for the Research

• 500 Hz Sampling Rate500 Hz Sampling Rate

• < $300 Project Expense< $300 Project Expense

• < 4 Degrees Accuracy< 4 Degrees Accuracy

Page 7: Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

Budget: ( Digi-Key, 1997 )

One Microchip 20 MHz microcontroller w/ 8 bit A/D ConverterPIC16C74A/JW-ND @ 18.77 each

$18.77One Microchip PIC16C7X Family Data Book

92678-ND @ $6.95 each $6.95

One CTS 20 MHz Microprocessor CrystalCTX062-N2 @ $1.74 each $1.74

One Power-One +5V Power SupplyPW9906-ND @ $109.72 each

$109.72One Burr-Brown CMOS 8 bit Buffered Digital-to-Analog Converter

DAC7528P-ND @ $6.30 $6.30

One Linear Technology Precision Voltage ComparatorLT1013ACN8-ND @ $7.13

$7.13One Burr-Brown Differential Precision Gain Amplifier

1NA105KU-ND @ $5.24 $5.24

8 – ¾” Rectangular / Multiturn ( 15 )Cermet/Industrial/Sealed Potentiometers3009P-103-ND @ $2.16 each

$17.2816 QT Optoelectronics Semiconductor Infrared Emitters

( IRED LED’s ) 1N6266QT-ND @ $4.15 each$66.40

16 QT Optoelectronics Semiconductor Infrared Photo DetectorsBPW38QT-ND @ $2.73 each

+ $43.68

Total $283.21

Page 8: Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

Initial BudgetInitial Budget• Microcontroller and instructionsMicrocontroller and instructions $ 25$ 25

• CrystalCrystal $ 2$ 2

• Power SupplyPower Supply $110$110

• ComparatorComparator $ 7$ 7

• Diff. OP-AMPDiff. OP-AMP $ 5$ 5

• PotentiometersPotentiometers $ 20$ 20

• IR EmittersIR Emitters $ 67$ 67

• IR Detectors IR Detectors $ 45$ 45

• TotalTotal $281$281

Page 9: Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

Targets for the ResearchTargets for the Research

• 500 Hz Sampling Rate500 Hz Sampling Rate

• < $300 Project Expense< $300 Project Expense

• < 4 Degrees Accuracy< 4 Degrees Accuracy

Page 10: Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

Why 500 Hz ?Why 500 Hz ?• SaccadesSaccades- rapid eye movements rapid eye movements

• last 30 - 120 ms, last 30 - 120 ms,

• cannot be within 100 ms of eachothercannot be within 100 ms of eachother

• videovideo eye tracking unit at 60 Hz eye tracking unit at 60 Hz (i.e. 17 ms, half a short saccade.) (i.e. 17 ms, half a short saccade.)

Page 11: Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

For a 20 MHz PIC16C74A/JW-ND processor

Sampling RateSampling Rate

• 200 200 sec / instructionsec / instruction

• 200 instructions200 instructions

• 40 40 sec for samplesec for sample

• 50 kHz50 kHz sample rate sample rate

Page 12: Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

For a 10 MHz PIC16C84 processor

Sampling RateSampling Rate

• 20 20 sec / instructionsec / instruction

• 100 instructions100 instructions

• 2 msec for sample2 msec for sample

• 500 Hz500 Hz sample rate sample rate

Page 13: Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

Targets for the ResearchTargets for the Research

• 500 Hz Sampling Rate500 Hz Sampling Rate

• < $300 Project Expense< $300 Project Expense

• < 4 Degrees Accuracy< 4 Degrees Accuracy

Page 14: Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

The The foveafovea covers approximately a 1 covers approximately a 100 Field of View. Video eye trackers can Field of View. Video eye trackers can come to within 1/2 come to within 1/2 0 0 for field of view. for field of view. The main consideration is speed.The main consideration is speed.

Why 4 Degrees?Why 4 Degrees?

Page 15: Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

To get within 4 Degrees:To get within 4 Degrees:

• individualindividual calibration calibration will be necessary will be necessary

• extensive extensive lookup tablelookup table creation creation

• consecutiveconsecutive ON/OFF ON/OFF state comparison state comparison ( provided for by software )( provided for by software )

Page 16: Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

Sample Calibration TargetSample Calibration Target

(with Nine Targets)(with Nine Targets)

Page 17: Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

Sample Calibration TargetSample Calibration Target

(with Seventeen Targets)(with Seventeen Targets)

Page 18: Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

The Proposed DesignThe Proposed Design

• ASL Model 210 Eye ASL Model 210 Eye Movement MonitorMovement Monitor

• made from scratchmade from scratch

• single eye single eye measurementmeasurement

• fit within HMDfit within HMD

Page 19: Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

HorizontalHorizontal++__

VerticalVertical | ? || ? |

ASL model 210 eye movement monitorASL model 210 eye movement monitor

Page 20: Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

Horizontal MeasurementHorizontal Measurement

++__

•Negative measurement = leftNegative measurement = left

•Positive measurement = rightPositive measurement = right

Page 21: Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

Detector Positions as the Detector Positions as the Viewer Shifts Their Viewer Shifts Their

Horizontal GazeHorizontal Gaze

RightRight CenterCenter LeftLeft

Page 22: Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

Vertical MeasurementVertical Measurement

| ? || ? |

•High measurement = upHigh measurement = up

•Low measurement = downLow measurement = down

Page 23: Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

Detector Positions as the Detector Positions as the Viewer Shifts Their Viewer Shifts Their

Vertical GazeVertical Gaze

UpUp CenterCenter DownDown

Page 24: Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

ASL video eye trackerASL video eye tracker

Page 25: Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

Long-term Goals of the Research

•Use in future projects involving Use in future projects involving eye eye contingent updatingcontingent updating and the study of and the study of perceptionperception..

• Measurement of Measurement of saccadicsaccadic eye eye movements.movements.

Page 26: Development of an Eye Tracker By Jason Kajon Barrett of the Center for Imaging Science at the Rochester Institute of Technology Advisor: Jeff Pelz.

ConclusionsConclusions

The eye tracker proposed is possible The eye tracker proposed is possible for the proposed for the proposed expenseexpense as there is as there is an adequate sacrifice for an adequate sacrifice for accuracyaccuracy of of eye position to increase the eye position to increase the speedspeed..

The resultant code written in The resultant code written in PicBasic will satisfy the design PicBasic will satisfy the design proposed. proposed.