1

Fracture Exercise• Laboratory experiments • Digital Imaging

Exercise (PBL-based)

• Compare images from different digital camerasand ”video” cameras.

• Demonstrate blooming/glare, low light conditionsand at least one more noise problem.

• Compare cameras and demonstrate at least onelimitation with a particular technology.

• Explain the physical reasons for the differencesyou observe.

Scene

Sensor

Observer

Imaging principles

Scene Lens Sensor

Image on film

Imaging principles

102030405060708090

100

200 400 600 800

sensitivity

(nm)UV IR

CCD ICX024Eye

Tungsten lamp2856 K

Night sky

Eyenight

Imaging principles Digital Imaging

2

FractureDetails

High Resolution Digital Camera

3120x2064

Pixels

14-bit (16384) resolution

8Mz readout700x500

16-bit (32768)

XMHz

LuLuminanceminance and and contrastcontrast

Scene luminance

Sensorsignal range

Monitordisplay range

1

10

100

1000

Black

White

verybright

verydark

dark

lightsaturation

noisyHigh

contrastscene

Lowcontrastscene

jai 75xjai

videoprocess

10

Scene illumination10

101010101101010-4

-3

-2

-1

1

2

3

4

5 Direct sunlightFull daylightOvercast dayVery dark dayTwilightDeep twilightFull moonQuater moonStarlightOvercast starlight

Dynamical resolution

High Resolution Digital Cameras

• Light sensitive• High spatial and

dynamical resolution• Low noise

MORE SENSITIVE THAN THE EYE

• Slow data transfer• Produces much data• Requires custom made

software• Not user friendly• Expensive

Advantages Drawbacks

3

CCD

CCD principle

CMOS

With on chip microlens

Interline Transfer CCD

Fill <100%Fill 35%

Photo diode

CMOS

Noise

• Shot noise / thermal / dark current• Read-out noise• Saturation / Glare / Blooming• High energetic ”cosmic” rays• ”Digital noise” / Moirè patterns

Binning

4

Frame Transfer and EEC

Frame-transfer CCD

Camera

Image

Scene

Process

Trans-port

Remote place Distance Localobserver

Video

1 FRAME = 2 FIELDSEven Odd

PAL 625x728 ”pixels” 50 Fields/25 Frames pr. Sec (528 image lines)

Video

Exposure TimeXSG

Video Out8 22

Selected exposure time

Video as result ofselected exposure time

Ver. sync

External Trigger

Exposure Time

WEN Out

Video Out

Hor. Sync

XSUB

XSG

Video

Video tape Frame grabber Color

5

400 500 600 700 800 900 1000Wave lenght (nm)

1.0

0.8

0.6

0.4

0.2

0.0

Rel

ativ

e re

spon

se

B G R

IR stopfilter

Color Bayer RGB filters

Blue pixel X,Y

Red pixel X,Y

Green pixel X,Y

3 chip CCD 3 chip CCD cameracamera

CoCo--site samplingsite sampling

RGB sensors arealligned pixel to pixel

FOV

PhotochargeEIA camera

Lamp supply50 Hz

Light

Shuttertime = 10 msec

Shutter time = one light period, photocharge = constant

Result = no flicker and reduced sensitivity

Flicker Optics

6

Test targets Image analysis

• Object properties

• Feature tracking / pattern matching

• Differences

• Transformations

Object Properties Object Properties

Object PropertiesSubSub pixel pixel measurementmeasurement

Fine line

Defocussedfine line

CL

Pixel signal

Imagedobject

Object center found by calculationswith pixel signals

7

Image analysis

• Object properties

• Feature tracking / pattern matching

• Differences

• Transformations

Velocity measurementsDV-images, 40ms apart, 1/16000 s exposure time

128x128 diamond shaped region tracking scheme

Tracking Scheme Explained Pixel Comparison AlgorithmsS1 = | If − I0 |

T∑

S2 = (If − I0)2

T∑

S3 =(If * I0)

T∑

( If2

T∑ Io

2

T∑ )

12

S4 =[(If − If) * (I0 − I0)

T∑ ]

( (If − If )2

T∑ (I0 − I0)2 )

12

T∑

Velocity fluctuationsTransformation

8

Differences

Differences Exercise (PBL-based)

• Compare images from different digital camerasand ”video” cameras.

• Demonstrate blooming/glare, low light conditionsand at least one more noise problem.

• Compare cameras and demonstrate at least onelimitation with a particular technology.

• Explain the physical reasons for the differencesyou observe.