Digital Imaging - BIOP -Homepagebiop.epfl.ch/pdf/DigitalImaging2010.pdf · Dr. Arne Seitz PT-BIOP...

Dr. Arne Seitz PT-BIOP course, Digital Imaging, EPFL 2010

BioImaging &Optics Platform

Dr. Arne SeitzSwiss Institute of Technology (EPFL)

Faculty of Life Sciences Head of BIOIMAGING AND OPTICS – BIOP

[email protected]

Digital Imaging



Digital imagingOutlook

1600 1700 1800 1900 2000

Digital Camera(Kodak, 1975)



Human Eye



Human Eye

• Visual Pigments– Protein + Retinal

– 13 cis/trans Photoreaction (Isomerization)

– Signal transduction pathway• Amplification steps

• Hyperpolarization

• Stimulated cells release less Neurotransmitter

– System is noiseless




Principle of digital imagingDifferent detection devices

– Photomultiplier– CCD Camera– etc.

Application in light microscopy– Do‘s and dont‘s– Pitfalls



What is lightparticle/wave

• 1669 NewtonEmanations theory (particle)

• 1677 Huygenswave theory

• 1802 Younginterference

• 1871 Maxwellelectromagnetic light theory

• 1886 Hertzexperimental proof of Maxwells theory



What is lightPhotoelectric effect

Albert EinsteinNobel price 1921

Experiments by Lennart 1902– Maximal energy of the photo

electrons is independent of the intensity of the light

– Slope is the same for different cathode materials

– Light can be described as particles called photons or light quants.

Detection of Photons



What is lightparticle/wave

Light can be described as– Electromagnetic wave

Amplitude, Phase, wavelength, polarization

– Particle (Photons)

Photoelectric effect can be used to detect photons.



Analog ImagingPhotographic plate

Detection Principle– Ag+ is reduced to Ag0

(complex photochemical reaction)

– development = amplification =more Ag+ is reduced to Ag0

– Fixationremoval of Ag+ ions

– NegativeLight exposed=blacknot exposed=transparent



Analog ImagingPhotographic plate

Advantages– Easy to handle.– High spatial resolution,

e.g. Fuji Velvia 50: 100 lines/mm.Disadvantages

– Low quantum efficiencyca. 0.8 % for photons butalmost 100% for electrons.

– “slow”, i.e. time consuming– Irreversible– Poor linearity





– Photomultiplier– Photodiode, CCD Camera– etc.

Application in light microscopy– Do‘s and don'ts– Pitfalls



Photomultiplier



PhotomultiplierSummary

• Photons “produce” electrons which are multiplied by acceleration and detected afterwards.

• Gain (=multiplication factor) can be varied.• Large linear range.• Quantum efficiency of the photocathode up to 30% in

the visible range.• “no” spatial resolution, i.e. photomultipliers can only be

used as point (scanning) detectors



Digital imagingDefinition

• A digital image is a representation of a two-dimensional image using ones and zeros (binary). (Wikipedia)

• Analog = continuous values• Digital = discrete steps

1

0

01

01

1 10 01

1 110 0

11 01 11 1



Digital imagingDefinition

• Array of photosensitive elements• Signal is dependent on the number of detected

photons; ideally linearly dependent=high dynamic range

• Easy read-out procedure• Reusable• Appropriate size, comparable to an analogue film



CCD CamerasPrinciple

E

conductor(metal)

Semi conductor Isolator

valenceband

conductionband

band gap



Semi conductor

E

CCD Cameras

Principle– Internal photoelectric

effect.

– Energy of the photon is used to transfer an electron from the valence band to the conduction band (semi conductors).

– Photodiodes

e- e- e- e-

e- e- e- e-p+ p+ p+ p+



CCD Architecture• Silicon based integrated

circuits• Dense matrix of

photodiodes• Electrons are stored in a

potential well• Electrons can be

transported across the chip (=read out)



CCD CameraBlooming



CCD CamerasBasic

Principle/Architecture

Types– Back illuminated, front illuminated

„Data“ Processing– Frame transfer, interline transfer, full frame



CCD CameraQuantum Efficiency



CCD Architecture



CCD Architecture

1

0

01

01

1 10 01

1 110 0

11 01 11 111 0

1

0

01

01

1 10 01

1 110 0

1 11 111 0

1

0

01

01

1 10 01

1 110 0

1 11 111 0

1

0

01

01

1 10 01

1 110 0

1 11



CCD Architecture



Spectral Detection

• Complementary metal oxide semiconductor (CMOS)

• Unacceptable performance until 1990

• Advantages•Low power consumption•Single voltage power supply





– Photomultiplier– Photodiode, CCD Camera– etc.

Application in light microscopy– Do‘s and don'ts– Pitfalls



Resolution/Pixel Size

Nyquist theorem

Undersampling/oversampling

Same areaLess pixels

Larger pixel size(in the image)

Less sampling frequency

Lateral resolution

NA61.0 λδ =R



Objective(numerical aperture)

ResolutionLimit

(microns)

ProjectedSize on CCD

(microns)

Required PixelSize

(microns)

4x (0.20) 1.5 5.8 2.910x (0.45) 0.64 6.4 3.220x (0.75) 0.39 7.7 3.940x (0.85) 0.34 13.6 6.840x (1.30) 0.22 8.9 4.560x (0.95) 0.31 18.3 9.260x (1.40) 0.21 12.4 6.2

100x (0.90) 0.32 32.0 16.0100x (1.25) 0.23 23.0 11.5100x (1.40) 0.21 21.0 10.5

Resolution/Pixel Size



CCD cameraBinning

Binning:- increases effective pixel size-Decreases sampling frequency-Increases read-out speed



CCD cameraNoise/Background

Noise=signal fluctuation

background

Sources of noise:•Detector noise

•Dark noise•Read noise

•Photon noise (shot noise)2Re

22

2

22

adDarkCam

Sig

CamSigtot

NNN

PQEN

NNN

+=

∗=

+=



CCD cameraSignal to Noise Ratio

Photonflux: 104 photons/s Exposure time: 0.1 sSNR: ~ 5 for a typical CCD camera



CCD cameraSignal to Noise Ratio



CCD cameraNoise



CCD CameraDynamic Range/Noise



CCD CameraDynamic Range



EM CCD



43

Readout noiseCamera Relative

gainAbsolute gain

Absolute noise, e-

Relative noise, e-

Camera offset

0 1 6.3 6.3 200CCD 128 5 33.2 6.6 200

255 10 54.8 5.5 2000 1 30.2 30.2 460

EMCCD 128 15 64.8 4.3 460255 550 887.9 1.6 460

Absolute readout noise increase with gain

Relative readout noise:CCD - does not depend on gainEMCCD - decrease with gain



44

SNR of CCD and EMCCD



CCD CameraSummary

• Photons “elevate” electrons from the valence to the conduction band

• High Quantum efficiency

• Large linear range.

• Spatial resolution is influenced by the pixel size of the camera

• EM CCD cameras for “low light” applications



More about digital imaging

1. Lecture Biomicroscopy I + II, Prof. Theo Lasser, EPFL

2. Interneta) http://micro.magnet.fsu.edub) b) Web sites of camera vendors, e.g.

• Hamamatsu• Rooper• Andor

•3. PT-BIOP•EPFL, SV-AI 0241, SV-AI 0140•http://biop.epfl.ch/

http://micro.magnet.fsu.edu/�

http://biop.epfl.ch/�

Digital Imaging - BIOP -Homepagebiop.epfl.ch/pdf/DigitalImaging2010.pdf · Dr. Arne Seitz PT-BIOP...

Documents

Transcript of Digital Imaging - BIOP -Homepagebiop.epfl.ch/pdf/DigitalImaging2010.pdf · Dr. Arne Seitz PT-BIOP...