Color - Dspace.mit.edu

ColorFrédo Durand and Barb CutlerMIT- EECSMany slides courtesy of Victor Ostromoukhov and Leonard McMillan

Why is the sky blue part II What do you mean exactly by blue?Color Vision2

Admin Quiz on Thursday– 1 sheet of notes allowed Review session tonight 7:30Color Vision3

Review of last weekColor Vision4

Monte Carlo Recap Random rays to sample rendering equation No meshing required, no special storage No limitation– On reflectance– On geometry Can be noisy (or slow Advanced1n)– Irradiance cache– Photon mapColor Vision5

You believe you know it all Color is about spectrumand wavelength We can get everythingfrom red, green and blue Well, life is moreconfusing than that!Color Vision6

Puzzles about color How comes a continuous spectrum ends upas a 3D color space Why is violet “close” to red Primaries: 3 or 4? Which ones– Red, blue, yellow, green– Cyan and magenta are not “spontaneous” primaries Color mixing What is the color of Henry IV’s white horse?Color Vision7

Plan Color VisionColor spacesProducing colorColor effectsColor Vision14

Cone spectral sensitivity Short, Medium and Long wavelengthS1.00M L0.750.500.250.00Color Vision400500 600wavelength70015

Cones do not “see” colorsM1.000.750.500.250.00Color Vision400500 600wavelength70016

Cones do not “see” colors Different wavelength, different intensity Same responseM1.000.750.500.250.00Color Vision400500 600wavelength70017

Response comparison Different wavelength, different intensity But different response for different conesS1.00M L0.750.500.250.00Color Vision400500 600wavelength70018

von Helmholtz 1859: Trichromatic theory Colors as relative responses(ratios)YellowOrangeRedShort wavelength receptorsRedOrangeYellowReceptor ResponsesGreenBlueVioletBlueGreenVioletMedium wavelength receptorsLong wavelength receptorsColor Vision400500600700Wavelengths (nm)19

Cones distribution In the retina LMS 40:20:1 No S (blue)in retina centerColor Vision20

Different spectrum Same responseINTENSITYMetamers400480 500580 600700INTENSITYWavelength (nm)400500600700Wavelength (nm)Color Vision21

Color matching Reproduce the color of atest lampwith the addition of 3primary lightsColor Vision22

Metamerism & light source Metamersunder a givenlight source May not bemetamersunder adifferentlampColor Vision23

Color blindness Dalton8% male, 0.6% femaleGeneticDichromate (2% male)– One type of cone missing– L (protanope), M (deuteranope),S (tritanope) Anomalous trichromat– Shifted sensitivityColor Vision24

Color blindness test Maze in subtle intensity contrast Visible only to color blinds Color contrast overrides intensity otherwiseColor Vision26

Plan Color Vision– Cone response, trichromats– Opponent theory– Higher-level Color spaces Producing color Color effectsColor Vision27

Remember von Helmholtz Colors as relative responses(ratios)YellowOrangeRedShort wavelength receptorsRedOrangeYellowReceptor ResponsesGreenBlueVioletBlueGreenVioletMedium wavelength receptorsLong wavelength receptorsColor Vision400500600700Wavelengths (nm)28

Hering 1874: Opponent Colors Hypothesis of 3 types of receptors: Red/Green, Blue/Yellow, Black/WhiteExplains well several visual phenomena 000---Color ptorsReceptors29

Dual Process Theory The input is LMS The output has a differentparameterization:– Light-dark– Blue-yellow– Red-greenMLSWhYRGBColor VisionTrichromaticOpponent-ProcessStageStageBk30

Color opponents wiring Sums for brightness Differences for color opponentsSML B Y-S-M-LColor Vision ML S-R G- M Y B- -L-M-ML W B-S M LL-S M LB W-G R---S-M-LM-L31

Simultaneous contrast In color opponentdirection Center-surroundColor VisionR G-B Y-G R-Y B-G R-B Y-R G-Y B-32

Land RetinexColor Vision33

Simultaneous Color ContrastColor Vision34

After-ImageColor Vision35

Opponent ColorsImageColor VisionAfterimage36

Opponents and image compression JPG, MPG Coloropponentsinstead ofRGB Compresscolor morethanluminanceColor Vision37

Plan Color Vision– Cone response, trichromats– Opponent theory– Higher-level Color spaces Producing color Color effectsColor Vision38

Color reparameterization The input is LMS The output has a differentparameterization:– Light-dark– Blue-yellow– Red-greenMLSWhYRB A later stage may reparameterize:– Brightness or Luminance or Value– Hue– SaturationColor VisionGBkL (or B)HS39

Hue Saturation ValueColor Vision40Courtesy of Stephen Palmer, VISION SCIENCE, and the MIT Press.Used with permission.

Hue Saturation Value One interpretationin spectrum space Not the only onebecause of metamerism Dominant wavelength(hue) Intensity Purity (saturation)Color Vision41

Color categories Prototypes Harder to classify colorsat boundariesDegree ofMembershipFocal GreenColor VisionFocal RedFocal BlueFocal YellowBlueYellow1(Red)0GreenRed(Blue)Hue42

Plan Color VisionColor spacesProducing colorColor effectsColor Vision43

Color spaces Human color perception is 3 dimensional How should we parameterize this 3D space Various constraints/goals––––Linear parameterizationClose to color technologyClose to human perceptionStandardColor Vision44

The root of all evil Cone responses are not orthogonal(they overlap) To change the M response without changing theL one, we need negative lightS1.00M L0.750.500.250.00Color Vision400500 600wavelength700Orthogonal basis (color matching function)45

Color Matching Problem Some colors cannot be produced using onlypositively weighted primaries E.g. primaries: pure wavelength– 650, 530, 460 Some colors need negative amounts ofprimaries Analysis spectrum hasnegative lobesColor Vision46

Color Matching Problem Some colors cannot beproduced using onlypositively weightedprimaries Solution: add light onthe other side!Color Vision47

Color Matching Problem Some colors cannot be produced using onlypositively weighted primaries Some tradeoff must be found betweennegative lobes in analysis vs. synthesis In 1931, the CIE(Commission Internationale de L’Eclairage)defined three new primaries Called X, Y , Z,– with positive colormatching functionsColor Vision48

CIE color space Can think of X, Y ,Z as coordinates Linear transform fromRGB or LMS(( ((( (RGBXYZ 3.24 -1.54 -0.50-0.97 1.88 0.040.06 -0.20 -1.06 0.410.210.020.36 0.180.72 0.070.12 0.95y(( ((( (XYZRGBxzColor Vision49

CIE color space Odd-shaped conecontains visible colors– Note that many pointsin XYZ do notcorrespond to visiblecolors!(( ((( (RGBXYZ 3.24 -1.54 -0.50-0.97 1.88 0.040.06 -0.20 -1.06 0.410.210.020.36 0.180.72 0.070.12 0.95y(( ((( (XYZRGBxzColor Vision50

CIE color space Objective, quantitative color descriptions– Dominant wavelength: Wavelength “seen” (corresponds to Hue)– Excitation purity: Saturation, expressed objectively– Luminance: Intensity Chromaticity (independent of luminance):– normalize against X Y Z:Color Vision51

CIE color space Spectrally pure colorslie along boundary Note that some huesdo not correspond toa pure spectrum(purple-violet) Standard white light(approximates sunlight)at 0.1490Blue480 Purple4000.1YellowCC580600Red7000.2 0.3 0.4 0.5 0.6 0.7 0.8xImage adapted from:Hunt, R. W. G. The Reproduction of Colour. John Wiley & Sons Incorporated. September2004. ISBN: 0-470-02425-9.Color Vision52

CIE color space Match color at somepoint A A is mix of whiteC, spectral B! What is dominantwavelength of A? What is excitationpurity (%) of A?– Move along AC/BCy0.80.7520540510B 0G0.2 0.3 0.4 0.5 0.6 0.7xImage adapted from:Hunt, R. W. G. The Reproduction of Colour. John Wiley & Sons Incorporated. September2004. ISBN: 0-470-02425-9.Color Vision53

XYZ vs. RGB Linear transform XYZ is more standardized XYZ can reproduce all colors with positivevalues XYZ is not realizable physically !!– What happens if you go “off” the diagram– In fact, the orthogonal (synthesis) basis of XYZrequires negative values.Color Vision54

Perceptually Uniform Space: MacAdam In color space CIE-XYZ, theperceived distance between colors isnot equal everywhereIn perceptually uniform color space,Euclidean distances reflectperceived differences betweencolorsMacAdam ellipses (areas ofunperceivable differences) becomecircles510490.05520530.10 .15540 550 560.65.60.55.50.45.25.30.20.25Spectrum locus570580590.35600610620.60650.40 .45 .50 .55.65 .70700 nm.20.15480yx470460.10.05plePu rline450 400 nmImage adapted from:Wyszecki, G. and W. S. Stiles. Color science: Concepts and Methods, Quantitative Dataand Formulae. Wiley-Interscience, 2nd ed. July 2000. ISBN: 0471399183.Color Vision55

CIE-LABImage adapted from Wyszecki, G., and W. S. Stiles,Color science: Concepts and methods, quantitative data and formulae.Wiley-Interscience; 2nd Edition. July 2000. ISBN: 0471399183Color VisionSource: [Wyszecki and Stiles ’82]56

Perceptually Uniform Space MunsellMunsell Color SpaceHueValueChromawww.munsell.comColor Visionmunsell.com57

Color response linear subspace Project the infinite-D spectrum onto a subspacedefined by 3 basis functions We can use 3x3 matrices to change thecolorspace– E.g. LMS to RGB– E.g. RGB to CIE XYZColor Vision58

Color response and RGB or LMS Project the infinite-D spectrum onto a subspace definedby 3 basis functions Small problem: this basis is NOT orthogonal What does orthogonal mean in our case? Second problem: the orthogonal basis is NOT physicallyrealizableColor Vision59

Color response and RGB or LMS Project the infinite-D spectrum onto a subspace definedby 3 basis functions Small problem: this basis is NOT orthogonal What does orthogonal mean in our case? Second problem: the orthogonal basis is NOT physicallyrealizableColor Vision60

Munsell book of colors Perceptually uniformwww.munsell.comColor Vision61

von Helmholtz 1859: Trichromatic theory Colors as relative responses (ratios) Violet Blue Green Yellow Orange Red Short wavelength receptors Medium wavelength receptors Long wavelength receptors Receptor Responses Wavelength