Human Colour Vision Linda Johansson, 2004
[email protected]
Contents INTRODUCTION A Brief History of Colour Vision Trichromatic Theory of Colour Vision Opponent-Process Theory of Colour Vision CIE System
3 3 4 4 5
LIGHT, THE VISUAL STIMULUS Electromagnetic Spectrum Photons Light Sources CIE Illuminants Light Material Interaction BidirectionalReflectance Distribution Functions
8 8 8 9 10 11 12
COLOUR STIMULUS Surface Reflectance Stimuli Metamerism Fluorescence
13 13 14 15 15
THE EYE Cornea Iris Lens Retina
16 16 16 17 17
VISUAL SIGNAL PROCESSING Signal Processing In the Retina Lateral Geniculate Nucleus (LGN) Visual Receiving Area
20 20 22 22
SENSITIVITY CONTROL Spectral Sensitivity of Rods and Cones Light and Dark Adaptation Chromatic Adaptation Dark Adaptation of Rods and Cones
23 23 24 25 25
LIGHTNESS AND COLOUR CONSTANCY Lightness Constancy Colour Constancy Chromatic Adaptation Memory Colours
27 27 27 27 28
SPATIAL AND TEMPORAL PROPERTIES OF COLOUR VISION Spatial and Temporal Frequency Contrast Sensitivity Functions (CSF) The Oblique Effect Mach Bands Flicker
29 29 30 31 31 31
COLOUR-VISION DEFICIENCY Monochromats Dichromats Anomalous Trichromats Trichromats Colour Vision Tests
32 32 32 33 33 33
SUBJECTIVE COLOUR PHENOMENA Simultaneous Contrast Crispening Spreading Luminance Phenomena Hue Phenomena Surround Phenomena
34 34 34 35 35 36 36
COLOUR ORDER SYSTEM The Munsell System NCS DIN OSA
37 37 38 39 40
TERMINOLOGY Colour Hue Brightness and Lightness Colourfulness and Chroma Saturation Related and Unrelated Colours Achromatic and Chromatic Colours
41 41 41 41 42 43 43 43
REFERENCES
44
2
1. Introduction The question of how, through the sense of vision, we are able to perceive colour of remote objects has been raised repeatedly throughout recorded history. Early philosophers and scientists held very different views regarding vision and colour perception than those now accepted in contemporary vision science.
1.1 A Brief History of Colour Vision Among the Greek philosophers it was widely believed that rays were discharged from the eyes (emanation theory) and that tiny replicas of perceived objects could be released by such rays, to be delivered through the pupil of the eye and from there flushed through the optic nerve to the sensorium in the brain. In the fifth century BCE, Empedocles (493-433BCE) wrote that the eye functioned like a lantern, that light from the eye shining outwards would interact with the “outer rays” and thereby allow objects to be seen. Aristotles (384-322 BCE) propagated a different notion of colour vision. He thought that colour was based on the interaction of stimulus brightness and ambient light level. He based this view on the perception that the colour of a sunset changed as darkness set in.
Alhazen (965-1040 CE)
In the Middle Ages the Arab scholar Alhazen (965-1040 CE) rejected the emanation theory. He correctly proposed that the eyes passively receive light reflected from objects, rather than emanating light rays themselves. He proposed a camera obscura model for the transmission of light in the eye, but did not speculate on the basis of colour vision. Leonardo da Vinci (1452–1519) came close to a full understanding of visual optics but was still convinced that the retinal image could not be inverted. The German astronomer Johannes Kepler (1571–1630) was the first to understand the basis of image formation by positive lenses and was, thereby, able to conclude that there must be an inverted retinal image. Sir Isaac Newton (1642–1727) demonstrated that the colours of objects relate to their spectral reflectance. He also stated correctly that the rays of light are not themselves coloured; rather they contain a disposition to elicit colour perceptions in an observer. It was Newton who gave the name spectrum to a strip of light shown through a prism and divided it into seven colours. The relationship between light and colour was revealed by Newton’s experiment. He also showed that the colours that compose white light could not be further subdivided but they could be recombined to form white light. His conclusion was that colour is not the product of the external objects we see, but is a property of the eye itself. This provided the foundation for modern theories of colour vision.
3
Kepler (1571-1630)
Newton (1642-1727)
1.2 Trichromatic Theory of Colour Vision In the later half of the nineteenth century, the trichromatic theory of colour vision was developed based on the work of James Clerk Maxwell (1831-1879), Thomas Young (1773-1829), and Hermann von Helmholtz (1821-1894). The trichromatic theory of colour vision proposes that colour vision depends on three receptor mechanisms, each with different spectral sensitivities [4][5]. The pattern of activity in the three mechanisms results in the perception of a colour. It was based on the results of a psychophysical procedure called colour matching. One of the more important empirical aspects of this theory is that it is possible to match all of the colours in the visible spectrum by appropriate mixing of three primary colours. Which primary colours are used is not critically important as long as mixing two of them do not produce the third. The trichromatic theory correctly explains one part of the colour vision process but the theory fails to explain several visually observed phenomena.
1.3 Opponent-Process Theory of Colour Vision The opponent-process theory of colour vision was proposed by Ewald Hering (1834-1918), who stated that colour vision is caused by opposing responses generated by blue and yellow, and by red and green [4]. It was based on the results of phenomenological observations involving afterimages, simultaneous contrast, colour visualization, and observations of the effect of colour blindness. These observations could not be accounted for by the trichromatic theory. For example, he noted that there are certain pairs of colours one never sees together at the same place and at the same time. For example, a colour perception is never described as reddish-green or yellowish blue. He also observed that there was a distinct pattern to the colour of the afterimages we see. For example that a red field generates a green afterimage and that viewing a green field generates a red afterimage, and that analogous results occur for blue and yellow. Hering also observed that people who are colourblind to red also are colour blind to green, and that people who can’t see blue also can’t see yellow. All these observations led to the conclusion that red and green are paired and that blue and yellow are paired [4][5]. It was popular in the first half of the 20th century for authors to pit the trichromatic theory against the opponent processes theory, but both the trichromatic and the opponent-process theories were proved to be correct. The reason for this is that the psycho-physical findings on which each theory was based were each reflecting physiological activity at different places in the visual system. The trichromatic theory operates at the receptor level and the opponent processes theory applies to the subsequent neural level of colour vision processing. The modern opponent-colour theory of colour vision explains how the two theories work together. The first stage of colour vision, the receptors, is indeed trichromatic, as hypothesized by Maxwell, Young, and Helmholtz [4][5]. However, contrary to simple thrichromatic theory, the three signals are not transmitted directly to the brain. Instead, the neurons of the retina encode the colour into opponent signals. The outputs of all the three cone types are summed (L+M+S) to produce an achromatic response and differencing of the cone signals allows construction of red-green (L–M+S) and yellow-blue (L+M–S) opponent signals [4]. The transformation from LMS signals to the opponent signals serves to
4
decorrelate the colour information carried in the three channels, thus allowing more efficient signal transmission and reducing difficulties with noise. The three opponent pathways also have distinct spatial and temporal characteristics that are important for predicting colour appearance.
1.4 CIE System Until 1931, there were no way to get a quantitative measurement description of colour and colours could only be specified by appeal to physical samples. In that year, the CIE (International Commission on Illumination) adopted a system of colour specification, which has lasted to present time. CIE has developed several colour systems based on a number of extensive measurements and experiments on how humans perceive colours. The sensitivity functions for the three cones (L, M, and S) were obtained through an experiment were an observer looked at a split screen with 100% reflection, i.e. a white surface. One half of the screen was illuminated by a reference light source, and the other half was illuminated by three light sources with red, green and blue light. The observers then tried to match the colour perceived from the reference light source with the colour perceived from the three monochromatic light sources by mixing them so that the two halves of the screen were perceived identical. The experiment was repeated with different reference wavelengths, but the same intensity and different observers. This is the simple form of colour matching and can be described by the following equation: Eq. 1.1:
M *P*w = M *t
where M is the measurement matrix, P is the primary spectra matrix, w=(w1…wk) is the weight vector and t=(t1…tN) is the test spectrum vector. Some of the reference colours could not be matched by any combination of the three primaries. In these cases, light from one or more of the primaries is added to the light of the reference colour. A match can then be achieved by adjusting the primaries in this configuration. Light that is added to the reference colour can be considered to have been subtracted from the mixture of the primaries. The advanced colour matching equation is given by: Eq. 1.2:
M * P * w1 = M * (t + P * q)
where w1 and q are the new weight vectors. The mean value from these tests constitutes the CIE colour matching functions, x(), y() and z(), which represents a standard observer, see Figure 1.1. From the colour matching functions, the tristimulus values, CIE XYZ can be calculated, see Eq. 1.3. These values are normalized for the current illumination so that a completely white surface always gives Y=100:
5
X = k R( )I( )x( )d
Y = k R( )I( )y( )d
Eq. 1.3:
Z = k R( )I( )z( )d
k=
100 I( )y( )d
I() is the spectral power distribution of the incident light and R( ) is the spectral reflectance of the object.
WAVELENGTH (nm)
Figure 1.1. CIE colour matching functions, x(), y() and z().
The colours can be represented in two dimensions by two chromaticity coordinates, x and y, that are independent of lightness (a definition of brightness and lightness can be found in 12.3), see equation 1.4. Eq. 1.4:
x=
X X +Y + Z
y=
Y X +Y + Z
These coordinates can be plotted in a chromaticity diagram, see figure 1.2. If all xy-coordinates for the pure wavelengths in the visible spectrum are plotted in the diagram, all will fall on a horseshoe shaped line, called the spectrum locus. The line that connects the end points of the spectrum locus is called the purple line. The colours on this line are a mixture of pure 380 nm (blue) and 770 nm (red) light. There is a white point in the middle of the chromaticity diagram where x=y=1/3.
6
550nm
500nm
White Point
600nm 700nm
Purple Line
x CHROMATICITY
Figure 1.2. CIE chromaticity diagram.
Although very useful, there are many limitations to this system. For one thing, the colour matches that it predicts apply only to a hypothetical standard observer, and not exactly to any particular human being. For another, it is valid only for restricted conditions of viewing with small fields that are neither too bright nor too dim. And, finally, the chromaticity diagram does not represent colour appearance very well, and although there is really no reason why it should, it has often been used for this purpose [5].
7
2. Light, the Visual Stimulus Light provides the electromagnetic energy required to initiate visual responses, i.e. it is the visual stimulus. Since the perception of colour begins with light, the colours that are perceived are influenced by the characteristics of the light source.
2.1 Electromagnetic Spectrum The receptors in our eyes are designed to receive and process electromagnetic energy from a very narrow band of energy within the electromagnetic spectrum that encompasses wavelengths between about 380 and 750 nm [5]. The wavelengths within this interval and their mixtures is called light, and light is the primary stimulus for colour vision. The energy in this spectrum can be described by its wavelength, i.e. the distance between the peaks of the electromagnetic waves. The wavelengths are associated with the different colours of the spectrum, see Figure 2.1.
400nm
500nm
600nm
700nm
Figure 2.1. Wavelengths and associated colours of the electromagnetic spectrum.
2.2 Photons Light consists of photons, which are indivisible units of radiant energy. The amount of energy associated with a photon of wavelength is: Eq. 2.1:
E=
hc
where E=energy, h=6.626x10-34J·s (Planck’s Constant), c=2.997x108m·s-1 (velocity of light), and =wavelength. The brighter a light is, the more photons are contained in it. Because photons are discrete packets of energy it is not possible to absorb a fraction of a photon. When a photon is emitted from a source it immediately moves at the speed of light [5]. If a photon moves with frequency v and in a plane perpendicular to its direction of travel at the speed of light c, then some distance will be traversed during the time required for the particle to move through one cycle. This distance is called wavelength and it is inversely proportional to the frequency: Eq. 2.2:
=
c
where c=2.997x108m·s-1(speed of light) and v=frequency.
8
2.3 Light Sources All visual perception requires a source of illumination that irradiates the objects that are seen. Because colour begins with light, the colours that are seen are influenced by the characteristics of the light source used for illumination. Two important concepts are Correlated Colour Temperature (CCT) and Colour Rendering Index (CRI). Colour temperature is a simplified way to characterize the spectral properties of a light source, while colour rendering index is a way to determine its quality. The Correlated Colour Temperature (CCT) of a light source is defined as the absolute temperature of a blackbody radiator (an “ideal”, hypothetical, body which absorbs all radiation falling on it) which produces the chromaticity nearest to that emitted by the light test source. It is measured in Kelvin (K) [4]. The CCT rating is an indication of how "warm" or "cool" the light source is. Low colour temperature implies warmer (more yellow/red) light while high colour temperature implies a colder (more blue) light. Some different colour temperatures and their corresponding colours are shown in Figure 2.2. 8000 K 7000 K 6000 K 5000 K 4000 K 3000 K 2000 K
Figure 2.2. Different colour temperatures and their corresponding colour appearance.
CRI (Colour Rendering Index) is a measure of how well the colours are reproduced by different light sources in comparison with a reference light source (typically a black body) at the same colour temperature. CIE have defined a method for how to determine CRI for light sources where the measure is graded from 1-100. Within this scale a CRI of 100 (optimal CRI) means that a sample illuminated with the light source is perceived to have the same colour as when illuminated with a reference light source. If a light source have a low CRI (50-60) it can cause severe colour distortions. It is preferable to have a light source with a CRI over 90 [2][6]. The natural illumination, sunlight, is the most important source of illumination. The solar energy is emitted from the sun and after interaction with the earth’s atmosphere it reaches the earth. Parameters such as solar elevation angle and atmospheric conditions will affect the overall intensity and spectral characteristics of direct solar illumination that, under normal conditions is the dominant source of illumination [5]. The production of artificial illumination originally required that something be burned in open air, for example the flame of a candle. Its spectral output in the shorter wavelengths of the visible spectrum is deficient, relative to that of daylight. In general, the same is true for incandescent lamps (ordinary light bulbs) in which a filament is heated until it glows [5]. When operated at very low current, no visible radiation is produced by such a lamp. As the applied voltage is increased,
9
causing an increase in current flow through the filament, its temperature is raised and the spectral distribution of the emitted light changes so that the level of short wavelength energy relative to long wavelength increases [5].
2.4 CIE Illuminants The CIE has established a number of spectral power distributions as CIE illuminants for colorimetry. These include CIE illuminants A, C, D65, D50, F2, F8, and F11. CIE Illuminant A is a mathematical representation of tungsten halogen (incandescent) having a colour temperature of 2 856 K. The colour of the light source is yellow/orange [4]. CIE Illuminant C is the spectral power distribution of illuminant A as modified by particular liquid filters defined by the CIE. It represents a daylight simulator with a CCT of 6 774 K. The colour of the light source is bluish [4]. CIE Illuminants D65 and D50 are part of the CIE D-series of illuminants that have been statistically defined based upon a large number of measurements of real daylight. Illuminant D65 represents an average daylight with a CCT of 6 504 K (neutral colour tone), and D50 (yellowish colour tone) represents an average daylight with a CCT of 5 003 K. D65 is commonly used in colorimetric applications, while D50 is often used in graphic arts applications [4].
Relative Spectral Power
CIE A
CIE D65
CIE C
CIE D50
CIE F11
Wavelength (nm)
Figure 2.3. Relative spectral power distributions for CIE illuminants A, C, D50, D65 and F11.
CIE F Illuminants (12 in all) represent typical spectral power distributions for various types of fluorescent sources. CIE illuminant F2 represents cool-white fluorescent with a CCT of 4 230 K. Illuminant F8 represents a fluorescent D50 simulator with a CCT of 5 000 K, and illuminant F11 represents a triband fluorescent source with a CCT of 4 000 K [4].
10
A light source can be described by the spectral power distribution, i.e. the power of its electromagnetic radiation as a function of wavelength or the number of photons as a function of wavelength. Figure 2.3 shows the spectral power distribution for two daylight illuminations (D50, D65), a tungsten lamp and a fluorescent lamp.
2.5 Light Material Interaction When light travels and encounters a medium other than that through which it has been travelling the light can be affected in many different ways. For example when light is incident upon a colour print, some of the light passes through the outer glossy surface and through the layers of selectively absorbing dyes. The spectral distribution of the light is altered by the double transfer through the dyes, both before and after reflection from the white paper surface beneath. There may also be internal reflections, which is a problem of scatter. Figure 2.4 illustrates the different paths a photon can take when it encounters a medium. Incident rays
Reflected ray
Air
Absorbed ray Scattered rays
Refracted ray
Transmitted ray
Figure 2.4. The various ways in which light rays interact when encountering a transparent medium.
Glossy surfaces reflect the light at the same angle of incidence and without any change of colour [5]. Whereas glossy surfaces are very smooth, matte surfaces have tiny surface imperfections that cause light to scatter and somehow have the ability to change colour. The most important property of a surface for perceiving its colour is diffuse spectral reflectance. This is a statement about how the probability of a photon being reflected from a surface, in an unpredictable direction, varies depending on the wavelength of the incident photon. When a beam of light enters some medium, not all of it will emerge out of the other side. Some of the light is absorbed by the medium and is converted to heat. The more transparent the medium is, the less absorption will take place. The extent to which absorption takes place is wavelength dependent [5]. Refraction refers to a change in the direction of light as it passes from one medium to another. Scatter occurs whenever the reradiation of photons by the molecules of transmitting medium is other than in the forward direction [5]. The light that we see when a beam pierces a smoky room is visible only because of scatter. When scattering particles are large, as they are in the eye, scatter is largely independent of wavelength and is concentrated in a
11
forward direction. When the particles are small, as in the atmosphere on a clear day; shortwave photons are much more likely to be scattered than long-wave ones; this is the physical basis for the blue colour of the sky. When light moves through a medium it is being transmitted. Except for vacuum, there is no such thing as a perfectly transparent medium. Transparent media have an effect on light because they contain atoms that interact with photons. If a light passes near the edge of a surface, it will appear to bend around the edge. This is called diffraction.
2.6 Bidirectional Reflectance Distribution Function (BRDF) Reflectance characteristics of objects can be described by a Bidirectional Reflectance Distribution Function (BRDF). This function describes what we all observe every day: that objects look differently when viewed from different angles, and when illuminated from different directions. The function describes the geometrical reflectance properties of a surface, i.e. how much light is reflected of a surface as a function of illumination geometry and viewing geometry at the light interaction point.
12
3. Colour Stimuli In colour science, a ”colour” that is to be viewed or measured is called more correctly a colour stimulus. This colour stimulus always consists of light. In some cases, that light might come directly from a light source itself, such as when a CRT screen or the flame of a lighted candle is viewed directly. But more typically, colour stimuli are the result of light that has been reflected from or transmitted through various objects.
3.1 Surface Reflectance How an object reacts to incident light depends upon various microscopic physical characteristics of its surface that determine the probability that an incident photon will be reflected in a particular direction depending upon its wavelength. It is important to distinguish clearly between two limiting aspects of surface reflection: diffuse, which mediates surface colour perception and specular, which usually does not [5]. An example of specular reflectance is provided by a dust-free mirror. If an ideal mirror’s edges are suitably disguised, its surface will be invisible. Light reflected from it does so at an angle equal to the angle of incidence, which is the geometrical property that allows a plane mirror to provide perfect virtual images. Objects are seen by reflection “in the mirror” as if they were located behind it; this happens because the complex, threedimensional flux of light reaching the eye is identical to what it would be if the perceived objects actually were located where they seem to be. Most surfaces exhibit reflectance components of both kinds. In very highly polished surfaces the outermost smooth layer of a hard surface can act as a mirror. But unlike a real mirror, such a surface is not totally reflecting. A significant fraction of the incident light penetrates the surface and is diffusely reflected by a substrate that contains dye and/or pigment particles collectively known as colorants. Diffuse reflection varies as a function of wavelength depending upon the nature of the colorants, and it is also affected to some extent by the type of binder that contains them [5]. Diffuse reflectance provides the physical basis for the colours of most objects. When light reaches an object, that light is absorbed, reflected, or transmitted. Depending on the chemical makeup of the object and certain other factors, the amount of light that is reflected or transmitted generally will vary at different wavelengths. This variation can be described physically by a spectral reflectance curve or a s p e c t r a l transmittance curve. These curves respectively describe the fraction of the incident power reflected or transmitted as a function of wavelength, see Figure 3.1 [3]. Note that the reflectance spectrum is has values between 0 and 1, while the illumination spectrum has values between 0 and .
13
Reflectance
Wavelength (nm)
Figure 3.1. Spectral reflectance for a red object.
In most cases, an object’s spectral characteristics will correlate in a straightforward way with the colour normally associated with the object. For example, the spectral reflectance shown in Figure 3.1 is for a red object. A red object (generally) is seen as red because it reflects a greater fraction of red light (longer visible wavelengths) than of green light (middle visible wavelengths) or blue light (shorter visible wavelengths). Sometimes, however, the correlation of colour and spectral reflectance is less obvious.
3.2 Stimuli
Wavelength (nm)
= Wavelength (nm)
Relative Power
X
Reflectance
Relative Power
If the object in Figure 3.1 is illuminated with the light source to the left in Figure 3.2 the colour stimulus will have the spectral power distribution shown to the right in Figure 3.2. The spectral power distribution of this stimulus is the product of the spectral power distribution of the light source and the spectral reflectance of the object. The spectral power distribution of the stimulus is calculated by multiplying the power of the light source times the reflectance of the object at each wavelength. For a reflective or transmissive object, the colour stimulus results from both the object and the light source. If another light source is used, the colour stimulus will change. A “red” object can be made to appear almost any colour, depending on how it is illuminated.
Wavelength (nm)
Figure 3.2. Calculation of the spectral distribution of a colour stimulus.
Note that the human eye is insensitive to light of wavelength greater than 650nm and less than 400nm [4]. This shows that even though the colour
14
stimuli of an object suggest one colour, the human perception may perceive a completely different colour, based on the sensitivity of the photoreceptors in the retina of the eye. The visual process is very complex and not fully understood. How a stimulus appears does not only depend on the spectral properties of the stimulus and the light source in which it is viewed. It also depends on many other factors like for example the size, shape and spatial properties and relationships of the stimulus, the background and surround, observer experience and the adapted state of the observer.
3.3 Metamerism Because of the trichromatic nature of the human vision, it is possible that two colour stimuli that are physically different (i.e., having different spectral power distribution) will appear identical to the human eye. This is called metamerism and two such stimuli are called a metameric pair [3]. The reason metamers look alike is that they both result in the same pattern of response in the three cone photoreceptors. Thus as far as the visual system is concerned, these stimuli are identical. In colour reproduction, metamerism is what makes colour encoding possible. It is because of metamerism that there is no need to reproduce the exact spectrum of a stimulus, rather it is sufficient to produce a stimulus that is a visually equivalent of the original one [3]. Note that, metamerism involves matching visual appearances of two colour stimuli, and not two objects. Hence, two different objects with different reflectance properties can form a metameric pair, under some special lighting conditions. Two stimuli that physically match, and for that reason also look identical are called isomers [5].
3.4 Fluorescence An important topic in colorimetric analysis of materials is fluorescence. Fluorescent materials absorb energy in one region of wavelengths and then re-emit this energy at another, usually longer, region of wavelengths [1]. For example, a fluorescent orange material might absorb blue photons and emit orange photons, i.e. some of the absorbed energy is emitted at longer wavelengths. A fluorescent material is characterized by its total radiance factor, which is the sum of the reflected and emitted energy at each wavelength relative to the energy that would be reflected by a PRD (Perfect Reflecting Diffuser) (A PRD is a theoretical material that is both a perfect reflector, i.e. it has 100% reflectance, and perfectly Lambertian, i.e. its radiance is equal in all directions) [1]. This definition allows for total radiance factors greater than 1.0, which is often the case. It is important to note that the total radiance factor will depend on the light source used in the measuring instrument, since the amount of emitted energy is directly proportional to the amount of absorbed energy in the excitation wavelengths. Spectrophotometric measurements of reflectance or transmittance of nonfluorescent materials are insensitive to the light source in the instrument, since its characteristics are normalized in the ratio calculations. This important difference highlights the major difficulty in measuring fluorescent materials [1].
15
4. The Eye Human vision is a complex process that involves the interaction of the two eyes and the brain through a network of neurons, receptors, and other specialized cells. The human eye is equipped with a variety of optical components including the cornea, iris, pupil, a variable-focus lens, and the retina. Together, these elements work to form images of the objects that fall into the field of view for each eye. When an object is observed, it is first focused through the cornea and lens, forming an inverted image on the surface of the retina, a multi-layered membrane that contains millions of light-sensitive cells that detect the image and translate it into a series of electrical signals for transmission via the optic nerves to the brain. In the brain, the optic nerves from both eyes join at the optic chaisma where information from their retinas is correlated. The visual information is then processed through several steps, eventually reaching the visual cortex, which is located on the lower rear section of each half of the cerebrum. Figure 4.1 shows a schematic representation of the optical structure of the human eye.
4.1 Cornea The cornea is the transparent outer surface of the front of the eye through which light enters the eye. It serves as the most significant image-forming element of the eye. This is because its refraction index (1.37) is substantially greater than that of air [5]. Thus, the smoothness of the corneal surface, and its index of refraction, are very important. Vision is unclear under water because water (1.33) and the cornea have nearly the same refractive index [5]. The optical power of the cornea is nearly lost and severe “farsightedness” (hyperopia) results [1].
4.2 Iris The iris is the muscle that controls the pupil size, and thus the amount of light entering the eye and reaching the retina. It is pigmented which gives the eye its specific colour. The size of the pupil depends mostly on the overall level of illumination, but it also depends on many other factors including the size and region of the retina stimulated, spectral and temporal characteristics of the light, and emotional reactions. Thus it is difficult to accurately predict the pupil size from the prevailing illumination. In practical situations, pupil diameter varies from about 3 RECEPTOR CELLS OPTIC NERVE FIBERS (RODS AND CONES) IRIS
PUPIL FOVEA CORNEA LENS
OPTIC NERVE RETINA
Figure 4.1. The eye [4].
16
CONE ROD RETINA PIGMENT EPITELEUM
mm to about 7 mm [1]. This change in pupil diameter results in approximately a five-fold change in pupil area and therefore in retinal illuminance. The pupils of both eyes change size together, called consensual papillary response, which means that both pupils grow smaller when light is delivered to only one of the eyes.
4.3 Lens The lens is a flexible structure with varying index of refraction. It is higher in the centre of the lens and lower at the edges. The shape of the lens is controlled by the ciliary muscles, and is called accommodation. When we gaze at a nearby object, the lens becomes “fatter” and thus has increased optical power to allow us to focus on the nearby object, see Figure 4.2. When we gaze at a distant object, the lens becomes “flatter”, thereby resulting in the decreased optical power required to bring far away objects into sharp focus. However, in some instances the components do not work correctly or the eye is slightly altered in shape and the focal point does not intersect with the retina. As people age, for instance, the lenses of their eyes become harder and loose their flexibility, which results in poor vision. If the point of an eye’s focus is short of the retina the condition is called nearsightedness or myopia. People with this affliction are unable to focus on distant objects. In cases where the eye’s focal point is behind the retina people have trouble focusing on nearby objects, which is a condition called hypermetropia, commonly known as farsightedness. These malfunctions of the eye can usually be corrected through the use of glasses, with concave lenses correcting myopia and convex lenses rectifying hypermetropia. The lenses can also become cloudy as one ages, called cataracts. DISTANT OBJECT
NEARBY OBJECT
NEARBY OBJECT
FOCUS ON RETINA
FOCUS ON RETINA
FOCUS BEHIND RETINA
Figure 4.2. Focusing Lens.
Concurrent with the hardening of the lens is an increase in its optical density. The lens absorbs and scatters short wavelength (blue and violet) energy. As it hardens, the level of this absorption and scattering increases. In other words, the lens becomes more and more yellow with age. Various mechanisms of chromatic adaption generally make us unaware of these gradual changes. However, we are all looking at the world through a yellow filter that not only changes with age but also significantly differs from observer to observer [1].
4.4 Retina The optical image formed by the eye is projected onto the retina. The retina is less than half a millimeter thick and contains a total area of about 1100 mm2. This area contains about 200 million neural cells that are directly involved with the processing of visual information [6].
17
The are two classes of photoreceptors in the human retina: rods and cones, see Figure 4.3, which transform light energy into electrical energy.
NUMBER OF RECEPTORS PER SQUARE MILLIMETER
Rods function under very low luminance levels (e.g., less than 1 cd/m2, where 1 cd correspond to the light emitted from one candle), while cones are used for high or daylight levels (e.g., greater than 100 cd/m2) and for seeing fine details [1]. BLIND SPOT RODS
RODS
CONES
CONES
ANGLE (DEG)
Figure 4.3. Rod and Cone distribution on the retina [4].
The rods are most sensitive to green wavelengths of light (about 510 nm), although they display a broad range of response throughout the visible spectrum [1]. Each eye contains about 120 million rods as compared to the number of cones that is only about 7 million [1]. The light sensitivity of rod cells is about 1000 times that of cone cells. However, the images generated by rod stimulation alone are relatively unsharp and confined to shades of grey. Rod vision is commonly referred to as scotopic vision. There are three types of cone receptors that absorb light from three different portions of the visible spectrum. The L-cones absorb longwavelength light (red), the M-cones absorb middle-wavelength light (green) and the S-cones absorb short-wavelength light (blue). Sometimes the cones are denoted with other symbols such as RGB or . The stimulation of the three types of cone receptors allows the human visual system to distinguish very small colour differences. It has been estimated that stimulation to various levels and ratios can give rise to about ten million distinguishable colour sensations. The relative distribution of the different cone types (L:M:S) on the retina is approximately 40:20:1 [1]. Stimulation of these visual receptors results in what is known as true colour vision. Cone vision is referred to as photopic vision. There is a difference in peak spectral sensitivity between scotopic and photopic vision. With scotopic vision, we are more sensitive to shorter wavelengths. This effect is known as the Purkinje shift and it can be observed by finding two objects, one blue and the other red, that appear the same lightness when viewed in daylight [1]. When the two objects are viewed under very low luminance levels, the blue object will appear quite light while the red object will appear nearly black because of the scotopic spectral sensitivity function. Near the centre of the retina is the area of sharpest vision, fovea centralis, that subtends about two degrees of visual angle (see section 8.1 for definition of visual angle). One method to measure the visual angle is the “Thumb method” where you fully extend your arm and look at your 18
thumb. The approximate visual angle of the thumb at arms length is 2 degrees [4]. The retina is less than half as thick in the fovea as in the remainder of the eye, and this change in thickness creates the depression from which the term fovea derives. The anatomy of the fovea pit has important implications for the resolution of fine visual detail. In order to see details (visual acuity) the eye needs to be focused on the fovea, which contains only high-density tightly packed cone cells. The density level of cone cells decreases outside of the fovea centralis and the ratio of rod cells to cone cells gradually increases. At the periphery of the retina, the total number of both types of light receptors decreases substantially, causing a dramatic loss of visual sensitivity. To have a retinal image of excellent optical quality formed upon the photoreceptors, it is necessary to reduce the scattered light within the retina as much as possible; this is neatly accomplished with the foveal depression. The improved spatial resolution that results is not accomplished at the expense of sensitivity to light. On the contrary, the fovea of the light-adapted retina is its region of highest sensitivity. There are no rods whatever in the central fovea. Located around 12°-15° from the fovea is the blind spot see Figure 4.3. This is the area where the optic nerve is formed and there is no room for photoreceptors.
19
5. Visual Signal Processing The following section explains the neural processing of visual information from the retina to the brain, i.e., the encoding of colour. Visual signal processing is a field of intense research and not all parts have been fully understood. All explanations given in the literature are considered to be of a more or less speculative nature. The following description of the process is somewhat simplified.
5.1 Signal Processing In the Retina Colour vision starts in the eye with the absorption of light by the outer segments of the photoreceptors, which contain visual pigment molecules that trigger electrical signals. These molecules have two components: a large protein called opsin and a small light-sensitive molecule called retinal [4]. Retinal, which is attached to the opsin reacts to light and is therefore responsible for the transformation of light energy into electrical energy (visual transduction). The transduction process begins when the light-sensitive retinal absorbs one photon of light. When the retinal absorbs this photon it changes its shape, a process called isomerization [4]. The electrical signals are then processed through a network of retinal neurons, which consists of four types of cells: bipolar cells, horizontal cells, amacrine cells, and ganglion cells, see Figure 5.1. RECEPTOR OUTER SEGMENTS RECEPTOR INNER SEGMENTS ROD AND CONE RECEPTORS (R)
RECEPTOR CELL BODIES
HORIZONTAL CELLS (H) BIPOLAR CELLS (B) AMACRINE CELLS (A)
GANGLION CELLS (G) OPTIC NERVE FIBERS
LIGHT Figure 5.1. The figure illustrates the signal processing from the photoreceptors to the ganglion cells [4].
20
The rods and cones connect differently to other neurons in the retina, they differ in the amount of convergence [4]. Convergence occurs when more than one neuron synapses on another neuron. In the retina 126 million receptors converge on 1 million ganglion cells [4]. Since there are 120 million rods but only 6 million cones, the rods must converge much more than the cones. On the average, about 120 rods pool their signals to one ganglion cell, but only about six cones send signals to a single ganglion cell. This difference between rod and cone convergence becomes even greater because of the fact that many of the foveal cones have “private lines” to ganglion cells. In these situations, with each ganglion cell receiving signals from only one cone, there is no convergence. The greater convergence of the rods compared to the cones translates into two differences in perception: the rods are more sensitive in the dark than the cones, and the cones result in better detail vision than the rods [4]. The receptors are connected to bipolar and horizontal cells by synapses. Together they form receptive fields where signals from a number of different photoreceptor cells are compared (Small receptive fields, i.e. fewer photoreceptor cells, provides greater visual acuity and large receptive fields, i.e. more photoreceptor cells, provides greater sensitivity). This causes an effect called center-surround antagonism. There are two basic types of bipolar cells: ON-center and OFF-center, see Figure 5.2. ON-center bipolar cells are activated by bright spots on dark surroundings, whereas OFF-cells are activated by black spots on light background.
–
+
+
–
ON-CENTRE BIPOLAR
–
+
OFF-CENTRE BIPOLAR
Figure 5.2. Receptive fields for bipolar cells: on-centre (left) and off-centre (right).
The center-surround receptive fields are sensitive to contrast. If both centre and surround are illuminated at the same time, the antagonistic effects almost cancel each other out. A consequence of this is that bipolars respond poorly to overall illumination levels, but are very sensitive to local differences in intensity, i.e. they are sensitive to contrast, not intensity. The impulses from the bipolar and horizontal cells are then transferred directly, or indirectly via amacrine cells, to the ganglion cells, that also have receptive fields with a centre-surround organization, just like the bipolar cells. It is the amacrine cells that add the surround signal to the ganglion cells and together they also form receptive fields. The ganglion cells have axons that leave the retina via the optic nerve, and connect them to the brain. The ganglion cells are of two major types: parvocellular ganglion cells (P cells) and magnocellular ganglion cells (M cells) that form two major parallel processing streams. P cells respond best to hight contrast, small objects (high spatial resolution) and slowly flashing stimuli (low temporal resolution). M cells respond best to the opposite, that is, low contrast, large objects (low spatial resolution) and fast flashing stimuli (high temporal resolution). The receptive fields of the M cells are also 21
larger compared to the P cells. Larger receptive field means more connections to photoreceptors and consequently the nerve impulses reach the brain more quickly.
RED ON
GREEN ON
YELLOW ON
BLUE ON
Figure 5.3. Colour opponent ganglion cells.
The final difference in behaviour between M and P cells is their response to light of different wavelengths. M cells are not wavelength selective and will respond to light of any colour. P cells do care about what colour the light is and are sensitive to wavelength in a ”colour opponent” way. Different types of colour opponent ganglion cells are shown in Figure 5.3. For example, they may have a centre that is excited by only green light (input from M-cones), and a surround that is inhibited by only red light (negative input from L-cones), see Figure 5.3 (left). Blue versus yellow centre-surround antagonism may also be found, see Figure 5.3 (right). P cells are the cells that form the basis of colour processing in the visual system.
5.2 Lateral Geniculate Nucleus (LGN) The optic nerve fibers enter the Lateral Geniculate Nucleus (LGN) in a layered structure with cells that respond to form, motion, and color. Here the process of co-ordinating vision from the two eyes starts. The LGN consists of six layers with each alternating layer receiving inputs from a different eye: 3 layers for the left eye and 3 layers for the right. The outer 4 layers (parvocellular layers) receive inputs from the P ganglion cells and the inner two layers (magnocellular layers) receive their input from the M ganglion cells. The result is three signals which are sent to the brain: one corresponding to the amount of green-or-red, one corresponding to the amount of blue-or-yellow, and one cooresponding to the lightness. The LGN cells then project to visual area one (V1) in the occipital lobe of the cortex. At this point, the information processing begins to become very complex.
5.3 Visual Receiving Area Most of the fibers from the LGN project to a region of the occipital cortex (outer layers of the brain at the back of the head) known as V1, primary visual cortex, or striate cortex. From V1 nerve fibres carry information to many other cortical areas. Approximately 30 visual areas have been defined in the cortex with names such as V2, V3, V4, and MT. The encoding of visual information becomes significantly more complex. Much as the outputs of various photoreceptors are combined and compared to produce ganglion cell responses, the outputs of various LGN cells are compared and combined to produce cortical responses. Beyond V1, there are two general streams of information processing: one for motion and location, and the other for colour and form. These are known as the ventral and dorsal streams, respectively. And in the end of this network of information, our ultimate perceptions are formed.
22
6. Sensitivity Control The human visual system is capable of functioning across vast changes in viewing conditions while providing relatively stable perceptions. The mechanism that allows the visual system to do this is known as adaptation. There are three types of adaptation: light, dark and chromatic. Light and dark adaptation describe the human visual system’s capability of functioning across large changes in luminance levels and chromatic adaptation is the ability of the human visual system to adjust to changes in the colour of illumination.
6.1 Spectral Sensitivity of Rods and Cones Perception is determined by the properties of the visual pigments. This can be shown by comparing the rods and cones spectral sensitivity, i.e., an observer’s sensitivity to light at each wavelength across the visible spectrum. The cone and rod spectral sensitivity curves are shown in Figure 6.1. The curves show that the rods are more sensitive to shortwavelength light than are the cones, with the rods being most sensitive to light of 500 nm and the cones being most sensitive to light of 560 nm [4]. The spectral sensitivity curves also show that the sensitivity of the human visual system rapidly decreases above 650 nm. That is why objects that have a high reflectance at longer visible wavelengths can appear having a certain colour even if the objects reflectance spectra tells otherwise. The human visual system also has very little sensitivity to wavelengths below 400 nm. This difference in the sensitivity of the rods and the cones to different wavelengths means that, as vision shifts from the cones to the rods during dark adaptation, we become relatively more sensitive to short-wavelength light, that is, light nearer the blue and green end of the spectrum [4]. The shift from cone vision to rod vision that causes the enhanced perception of short wavelengths during dark adaptation is called the Purkinje shift, after Johann Purkinje, who described this effect. ROD VISION
BLUE
CONE VISION
GREEN
YELLOW
RED
Figure 6.1. Spectral sensitivity for rod vision and cone vision [4].
The difference between the rods and cone spectral sensitivity curves is caused by differences in the absorption spectra of the rod and cone visual pigments [4]. The absorption spectra of the rod and cone pigment are shown in Figure 6.2. 23
CONES
RELATIVE PROPORTION OF LIGHT ABSORBED
ROD
WAVELENGTH (NM)
Figure 6.2. Absorption spectra of the human rod pigment and the three human cone pigments [4].
The rod pigment absorbs best at 500 nm, the blue-green area of the spectrum. There are three absorption spectra for the cones because there are three different cone pigments, each contained in its own receptor. The short-wavelength pigment absorbs light best at about 419 nm; the medium-wavelength pigment absorbs light best at about 531 nm; and the long-wavelength pigment absorbs light best at about 558 nm [4]. The absorption of the rod visual pigment closely matches the rod spectral sensitivity curve, and the short-, medium-, and long-wavelength cone pigments add together to result in a psychophysical spectral sensitivity curve that peaks at 560 nm [4]. Since there are fewer short-wavelength receptors and therefore much less of the short-wavelength pigment, the spectral sensitivity curve is determined mainly by the medium- and longwavelength pigments [4]. It is clear that the rates of rod and cone dark adaptation and the shapes of the rod and cone spectral sensitivity curves are determined by the properties of the rod and cone visual pigments.
6.2 Light and Dark Adaptation Light adaptation is the decrease in visual sensitivity as a function of the overall amount of illumination [1]. The more light illuminating a scene, the less sensitive the human visual system becomes to light. Dark adaptation is the opposite of light adaptation, i.e., the change in visual sensitivity that occurs when prevailing level of illumination is decreased, opposite to light adaptation. The human visual system becomes more sensitive to light as the overall amount of illumination decreases. This can be thought of as walking from the sunny afternoon light into a darkened room. After several minutes, objects become recognizable as your visual system adapts. The visual sensitivity will gradually improve and eventually (in about 30 minutes) reach a state that is optimal for that amount of illumination. This happens because the visual system is responding to the lack of illumination by becoming more sensitive and therefore capable of producing s meaningful visual response at the lower illumination level. Light and dark adaptation function at different speeds. The speed of adaptation is called the time-course for full adaptation. Light adaptation works at a much faster rate than dark adaptation, being on the order of 5 minutes compared to 30 minutes for dark adaptation [4].
24
6.3 Chromatic Adaptation Chromatic adaptation refers to the human visual system’s ability to adjust to the colour of overall illumination rather than the absolute levels of the illumination. Consider a white object such as a piece of white paper. This paper can be viewed under a variety of light sources such as daylight, incandescent, and fluorescent. Despite the large change in the colour of these sources (ranging from blue to orange), the paper will always retain an approximate white appearance. This is because the Scone system becomes relatively less sensitive under daylight to compensate for the additional short-wavelength energy, while the L-cone system becomes relatively less sensitive under incandescent illumination to compensate for the additional long-wavelength energy.
6.4 Dark Adaptation of Rods and Cones Dark adaptation is a two-stage process where the increased light sensitivity takes place in two distinct stages: an initial rapid stage and a later, slower stage. Figure 6.3 shows the dark adaptation curves (the eye’s light sensitivity over time). The curves indicate that the observer’s sensitivity increases in two phases. It increases rapidly for the first 3 to 4 minutes after the light is extinguished and then levels off; then, after about 7 to 10 minutes, sensitivity begins to increase further and continues to do so for another 20 to 30 minutes [4]. The sensitivity at the end of dark adaptation, labelled dark-adapted sensitivity, is about 100000 times greater than the light-adapted sensitivity measured before dark adaptation began [4]. The initial rapid stage is due to adaptation of the cone receptors and the second slower stage is due to adaptation of the rod receptors. ROD LIGHT-ADAPTED SENSITIVITY
LOW ROD
CONE LIGHT-ADAPTED SENSITIVITY
ROD-CONE BREAK MAXIMUM CONE SENSITIVITY
CONE
DARK-ADAPTED SENSITIVITY HIGH
MAXIMUM ROD SENSITIVITY TIME IN DARK (MIN)
Figure 6.3. Dark-adaptation curves [4].
25
Both the rods and cones begin gaining in sensitivity as soon as the lights are extinguished, but since the cones are more sensitive at the beginning of dark adaptation, they determine the early part of the dark-adaptation curve. After about 3 to 5 minutes, the cones finish their adaptation, and the curve levels off. However, by about 7 minutes after the beginning of dark adaptation, the rods finally catch up to the cones and then become more sensitive. When this occurs, the curve starts down again, creating the rod-cone break, which is where the sensitivity of the rods begins to determine the dark adaptation curve. As the rods continue their adaptation, the dark adaptation curve continues downward for about 15 more minutes. The rods reach their maximum sensitivity about 20 to 30 minutes from the beginning of dark adaptation, compared to only 3 to 4 minutes for the cones [4]. These differences in the rate of adaptation can be traced to a process called visual pigment regeneration that occurs with different speeds in the rods and the cones [4]. When the visual pigment absorbs light, the lightsensitive retinal molecule changes shape and triggers the transduction process. It then separates from the larger opsin molecule, and this separation causes the retina to become lighter in colour, a process called pigment bleaching [4]. Before the visual pigment can again change light energy into electrical energy, the retinal and the opsin must be rejoined. This process, which is called pigment regeneration, occurs in the dark with the aid of enzymes supplied to the visual pigments by the nearby pigment epithelium. As the retinal and opsin components of the visual pigment recombine in the dark, the pigment begins to become darker again.
26
7. Lightness and Colour Constancy Lightness and colour constancy helps to keep our perception of achromatic and chromatic colours constant even when the illumination changes. This means that we can perceive the actual properties of objects without too much interference from different light sources.
7.1 Lightness Constancy Lightness constancy refers to how our perception of lightness remains relatively constant even when objects are viewed under different intensities of light [4]. If a brighter light hits an object more light hits the object and therefore much more light is also reflected. But the perception of the shade of lightness remains the same regardless of the changes in the amount of light reflected into the eyes. Our perception of an object’s lightness is related not to the amount of light that is reflected from the object, which can change depending on the illumination. But on the percentage of light reflected from the object, which remains the same no matter what the illumination. Objects that look black reflect about 5 percent of the light, objects that look grey reflect about 10 to 70 percent of the light, and objects that look white reflect about 80 to 90 percent of the light [4].
7.2 Colour Constancy Colour constancy refers to how our perception of colour remains relatively constant even when objects are viewed under different illuminations [4]. As an example the colour of objects do not change when moving from indoors to outdoors, even though the illumination condition has changed dramatically. Figure 2.3 showed the wavelengths that are contained in light from a lightbulb (CIE Illuminant A) and the wavelengths contained in sunlight (CIE Illuminant D50 and D65). The sunlight contains approximately equal amounts of energy at all wavelengths, which is a characteristic of white light. The bulb contains much more energy at long wavelengths. Even though there is a big difference between the wavelength distribution of the sunlight and the lightbulb, we do not notice much change in how we perceive the colours of objects under these two different light sources. Although small shifts of colour perception sometimes occur when the illumination changes, our overwhelming experience is that colours remain at least approximately constant under most natural conditions. Colour constancy is due to a number of factors including chromatic adaption, the effect of surrounds, and memory colour.
7.3 Chromatic Adaptation One of the mechanisms that contributes to colour constancy is chromatic adaptation, i.e. prolonged exposure to a chromatic colour. When we walk into a room illuminated with a tungsten light, the eye adapts to the long-
27
wavelength-rich tungsten light, which decreases the eye’s sensitivity to long wavelengths. This decreased sensitivity causes the long-wavelength light reflected from objects to have less effect than before the adaptation, and this compensates for the greater amount of long-wavelength tungsten light that is reflected from everything in the room. The result is just a small change in the perception of colour. The eye is adjusting its sensitivity to different wavelengths in order to keep colour perception approximately constant under different illuminations [4].
7.4 Memory Colours An object’s perceived colour is affected not only by the observer’s state of adaptation. Another small effect is that past knowledge can have some effect on colour perception through the operation of a phenomenon called memory colour, in which an objetc’s characteristic colour influences our perception of its colour. Research has shown that since people know the colours of familiar objects, like a red stop sign, or a green tree, they judge these familiar objects as having richer, more saturated colours than unfamiliar objects that reflect the same wavelengths [4]. Thus, our ability to remember the colours of familiar objects may help us perceive these colours under different illuminations.
28
8. Spatial and Temporal Properties of Colour Vision Our eyes are constantly sampling information of images projected onto the retina. Information is then integrated so objects around us (overall shapes and small details) appear clearly visible and also appear to be stable or move smoothly. Since there is a finite amount of field of view and time required to collect and process information, there are limitations to the responsiveness of our visual system to details and rates of change.
8.1 Spatial and Temporal Frequency Spatial frequency is how rapidly a stimulus changes across space, with high spatial frequencies corresponding to small details in the environment and low spatial frequencies corresponding to larger forms [4]. For example, the grating to the right in Figure 8.1 has a higher spatial frequency than the one to the left, because it has more bars per unit distance.
Figure 8.1. Two gratings. The one to the right has a higher spatial frequency than the one to the left.
Spatial frequency is measured in terms of cycles per degree of visual angle. The visual angle is the angle of an object relative to the observer’s eye, see figure 8.2. The visual angle depends on both the size of the stimulus and on its distance from the observer. If the distance is increased, the visual angle becomes smaller. The term “cycles per degree” means the number of cycles in a grating that fit within an angle of one degree on the retina, where one cycle is a dark bar and a light bar. VISUAL ANGLE RETINAL IMAGE
Figure 8.2. Visual angle, the angle of an object relative to the observer’s eye.
The experimental procedure used in studying the spatial characteristics of the visual system typically involves a visual stimulus that is displayed in the form of a sine-wave grating, that is, a regular stripe pattern whose luminance across the pattern varies sinusodially [4]. The observer is asked to determine the threshold for detecting the pattern. Presented with a sine-wave grating of given spatial frequency, the observer adjusts the amplitude of the luminance variation until he or she can just see the presence of the grating, or just distinguish it from a perfectly uniform field.
29
In the temporal domain, the same principle applies, except now the stimulus is separated in time, i.e. temporal frequency is how rapidly a stimulus is changes over time. Temporal frequency is measured in Hz. The experimental procedure used in studying the temporal response characteristics of the visual system typically involves a visual stimulus of some specified size whose luminance is varied sinusoidally as a function of time over a range of frequencies and luminance amplitudes. For a given frequency and mean luminance, the observer adjusts the luminance amplitude until the imposed sinusoidal variation is just large enough that the field does not appear steady in brightness.
8.2 Contrast Sensitivity Functions (CSF) The spatial and temporal characteristics of the human visual system can be measured as so-called contrast sensitivity functions (CSF). A CSF is a plot of contrast sensitivity vs. spatial or temporal frequency [1]. Contrast is typically defined as the difference between maximum and minimum luminance in a stimulus divided by the sum of the maximum and minimum luminances, and CSFs are typically measured with stimuli that vary sinusodially across space or time. Figure 8.3 (left) illustrates typical spatial CSFs for luminance (blackwhite) contrast and chromatic (red-green and yellow-blue at constant luminance) contrast. The luminance CSF has band-pass characteristics, with a peak-sensitivity around 5 cycles per degree. This function approaches zero at zero cycles per degree, thus illustrating the tendency for the visual system to be insensitive to uniform fields. It also approaches zero at about 60 cycles per degree, the point at which detail can no longer be resolved by the eye. The band-pass CSF correlates with the concept of center-surround antagonistic receptive fields that would be most sensitive to an intermediate range of spatial frequency. The chromatic mechanisms have low-pass characteristics and have significantly lower cutoff frequencies. This indicates the reduced availability of chromatic information for fine details (high spectral frequencies) that is often taken advantage of in image coding and compression schemes (e.g. JPEG). The low-pass characteristic of the chromatic mechanisms also illustrate that edge detection/enhancement does not occur along these dimensions. The blue-yellow chromatic CSF has a lower cutoff frequency than does the red-green chromatic CSF due to the scarcity of S cones in the retina. The luminance CSF is significantly higher than the chromatic CSFs. This indicates that the visual system is more sensitive to small changes in luminance contrast compared to chromatic contrast. Figure 8.3 (right) illustrates typical temporal CSFs for luminance and chromatic contrast. They share many characteristics with the spatial CSFs. Again, the luminance temporal CSF is higher in both sensitivity and cutoff frequency (close to 60 Hz) than are the chromatic temporal CSFs. Also, it shows band-pass characteristics that suggest the enhancement of temporal transients in the human visual system. The spatial and temporal CSFs interact with one another. A spatial CSF measured at different temporal frequencies will vary tremendously, as will a temporal CSF measured at various spatial frequencies. Many visual scientists have directed their attention to visual phenomena that are mainly associated with temporal and spatial variations of the
30
CONTRAST SENSITIVITY
CONTRAST SENSITIVITY
LUMINANCE
RED-GREEN
LUMINANCE
CHROMATIC
BLUE-YELLOW
LOG TEMPORAL FREQUENCY (Hz)
LOG SPATIAL FREQUENCY (cpd)
Figure 8.3. Spatial contrast sensitivity functions (left) and temporal contrast sensitivity functions (right) for luminance and chromatic contrast [1].
observed stimuli. In particular, the objective has been to quantify the temporal and spatial response characteristics of the visual system. The tool of Fourier analysis have been applied effectively to this task yielding temporal and spatial modulation transfer functions of the visual system.
8.3 The Oblique Effect Humans are more sensitive to horizontally or vertically oriented gratings than to other, oblique, orientations. This enhanced sensitivity for vertical and horizontal gratings is called the oblique effect. This phenomenon is considered in the design of rotated halftone screens that are set up such that the most visible pattern is oriented at 45°.
8.4 Mach Bands Mach Bands are light or dark narrow bands that are perceived near the border of two adjacent fields, one field being darker than the other, see Figure 8.4. Between two regions of different intensity a thin bright band appears at the lighter side and a thin dark band appears on the darker side. These bands are not physically present they are just illusions.
Figure 8.4. Mach Band.
The phenomenon in all its complexities is not fully understood but it is generally agreed that lateral interactions in the neural network of the visual system account for it.
8.5 Flicker When intermittent stimuli are presented to the eye they are perceived as separate if the rate at which they are presented is below a certain value. Depending on the speed the intermittent stimulation of the observer’s visual system results in the sensation of flicker. At a very slow rate it appears to flash on and off in a discrete but regular fashion. If the rate is increased, then, above a certain critical rate, the flicker ceases. This point is called the critical flicker frequency (CFF) and is influenced by a number of factors. The phenomenon of the disappearance of flicker at that frequency is called flicker fusion [6]. 31
9. Colour-Vision Deficiency Colour-vision deficiency is an inability to perceive some of the colours that people with normal colour vision can perceive. People with colour deficiency (dichromats) and colour blindness (monochromats) need fewer wavelengths than a normal trichromat to match any wavelength in the spectrum. There are three types of colour deficiency: monochromats who need only one wavelength to match any colour in the spectrum, dichromats who need only two wavelengths to match all other wavelengths in the spectrum and anomalous trichromats who need three wavelengths to match any wavelength, just as a normal trichromat does, but the anomalous trichromat mixes these wavelengths in different proportions from a trichromat. An anomalous trichromat also have difficulties in discriminating between wavelengths that are close together. Colour-vision deficiencies are not rare, particularly in the male population where about 8% have some type of colour-vision deficiency as compared to the female population where the number is only 0.4% [1]. The reason for this disparity is genetic. The genes for photopigments are present on the X chromosome. Since males (XY) have only one X chromosome, a defect in the visual pigment gene on this chromosome causes colour deficiency. Females (XX) with their two X chromosomes are less likely to become colour deficient, since only one normal gene is required for normal colour vision. If a female is colour-deficient, it means she has two deficient X chromosomes and all male children are destined to have colour-vision deficiency [1][4].
9.1 Monochromats People with no functioning cones (i.e., only rod vision in both dim and bright light) are called rod monochromats or achromats. Mono-chromats see everything in shades of lightness and can therefore be called colourblind. Only 0.001% of the population are monochromats and it is hereditary [4]. Another group of people who are truly colour-blind are those that have rods and only one class of cone receptors. At photopic levels such observers would not be able to distinguish one colour from another. These observers are called cone monochromats.
9.2 Dichromats Those who have two classes of functioning cones are called dichromats. Dichromats experience some colours, though a lesser range than trichromats. There are three different forms of dichromats depending on which one of the three normal photopigments (L, M, S) is missing. An observer with tritanopia (0.002% of males, 0.001% of females) is missing the S-cone photopigment and therefore cannot discriminate yellowish and bluish hues [4]. A deuteranope (1% of males, 0.01% of females) is missing the M-cone photopigment and therefore cannot distinguish reddish from greenish hues [4]. And a protanope (1% of males, 0.02% of females) is missing the L-cone photopigment and therefore is also unable to discriminate reddish and greenish hues [4]. 32
9.3 Anomalous Trichromats Observers who have three classes of cones but don’t see the world as socalled colour normal observers are called anomalous trichromats (abnormal trichromatic vision). In this case, the ability to discriminate particular hues is reduced either due to shifts in the spectral sensitivities of the photopigments or the contamination of photopigments (e.g., some L-cone photopigment in the M-cones, and so on) [1]. Among the anomalous trichromats are those with any of the following: protanomaly, that is, either they are weak in L-cone photopigment or the L-cone absorption is shifted toward shorter wavelengths, deuteranomaly, that is, either they are weak in M-cone photopigment or the M-cone absorption is shifted toward longer wavelengths and tritanomaly, that is, either they are weak in S-cone photopigment or the S-cone absorption is shifted toward longer wavelengths [1].
9.4 Trichromats There are also colour vision variations among observers with normal colour vision, trichromats. There can for example be differences in the proportion of the different cone types or variances in the peak spectral absorbance of the cone photopigments.
9.5 Colour Vision Tests There are many different types of colour vision tests available. Some tests are very quick and makes it possible to differentiate colour normals from those who clearly have a colour vision deficiency in just a few minutes, while other tests takes considerably longer to administer. One of the well-known quick tests uses the Ishihara Plates, which belong to a category of tests called pseudoisochromatic plates. Another common test that measures the observer’s ability to make very subtle colour discrimination is the Farnsworth-Munsell 100 Hue test. Pseudoisochromatic plates are colour plates made up of dots of various colours. The test takes advantage of one of the Gestalt laws of organization, the law of similarity. According to this principle, elements having the same appearance tend to be apprehended as a pattern. By manipulating the chromaticities of such elements at constant luminance, they can form a figure and a background. The plates are presented under properly controlled illumination to observers who are asked to respond by either tracing the pattern or reporting the number observed. Various plates are designed with colour combinations that would be difficult to discriminate for observers with the different types of colour-vision deficiencies. The Farnsworth-Munsell 100-Hue Test consists of four sets of chips that must be arranged in an orderly progression of hue. Observers with various types of colour-vision deficiencies will make errors in the arrangement of the chips at various locations around the hue circle. The test can be used to distinguish between the different types of deficiencies and also to evaluate the severity of colour discrimination problems. It also can be used to identify observers who have normal colour vision but poor colour discrimination for all colours.
33
10. Subjective Colour Phenomena Colour vision is a very complex phenomenon and the colour of an object depends not only on the nature of the paint on its surface, but also on the colour of the light used to illuminate it, the intensity of that light, and the chromatic characteristics of other surfaces located nearby. The colours we perceive can be separated into objective colours and subjective colours. Objective colours are the perception of colour consistent with what is expected in response to a particular spectral distribution of energy. In contrast, subjective colours are produced within the visual system without being related directly to specific wavelengths of light. Objective colours are initiated by the differential activation of the three kinds of cone photoreceptors. Whereas subjective colours are seen when this initial receptor stage of vision is bypassed. Here follows some examples of subjective colours.
10.1 Simultaneous Contrast Simultaneous contrast causes a stimulus to shift in colour appearance when the colour of its background changes. A light background induces a stimulus to appear darker, a dark background induces a lighter appearance, red induces green, green induces red, yellow induces blue, and blue induces yellow, see Figure 10.1.
Figure 10.1. An example of simultaneous contrast. All the grey patches to the left are physically identical, and all the red and green patches are identical.
10.2 Crispening Crispening is the increase in perceived colour difference between two stimuli when the background of the stimuli is close to the colour of the stimuli itself. The figure below illustrates crispening for a pair of grey samples. The two grey stimuli appear to be of greater lightness difference on the grey background than on either the white or the black background, see Figure 10.2.
34
Figure 10.2. An example of crispening. The pairs of grey patches are physically identical on all three backgrounds.
10.3 Spreading Spreading is the mixture of a colour stimulus with its surround. When the stimuli increase in spatial frequency, or become smaller, the simultaneous contrast effect disappears and is replaced with a spreading effect, see Figure 10.3.
Figure 10.3. An example of spreading.
The effect works as if it is combining the adjacent colours rather than accentuating the differences between adjacent colours as contrast does.
10.4 Luminance Phenomena Hunt effect (Colourfulness increases with luminance) – As the luminance of a given colour increases, its perceived colourfulness also increases [1]. Objects appear much more vivid, or colourful, when viewed in bright sunny environment. Stevens effect (Contrast increases with luminance) – As the luminance level increases, so too does the brightness contrast [1]. As the adapting luminance level increases, the rate of change between the brightness of the dark and light colour increases. This rate of change is often considered to be the contrast of the scene. Helmholtz-Kolrausch effect (Brightness depends on luminance and chromaticity) – Brightness changes as a function of saturation, i.e., as a stimulus becomes more saturated at constant luminance, its perceived
35
brightness also increases [1]. A chromatic stimulus will appear brighter than an achromatic stimulus at the same luminance level.
10.5 Hue Phenomena Bezold-Brücke hue shift (Hue changes with luminance) – Illustrates that the wavelength of monochromatic light sources is not a good indicator of perceived hue [1]. As luminance levels change the perceived hue can also change. Abney effect (Hue changes with colorimetric purity) – States that adding ”white” light to a mono-chromatic light does not preserve constant hue [1]. Straight lines in a chromaticity diagram radiating from the chromaticity of the white point to the spectral locus, are not lines of constant hue. Unlike the Bezold-Brücke hue shift, this effect is valid for related as well as unrelated colours. Helson-Judd effect (Hue of nonselective samples) – Illustrates that nonselective (grey) stimuli viewed under highly chromatic illumination take on the hue of the light source if they are lighter than the background, and they take on the complementary hue if they are darker than the background [1].
10.6 Surround Phenomena Bartleson-Braneman Equations (Image contrast changes with surround) – Perceived contrast in images increases as the luminance of the surround increases [1]. When an image is viewed in a dark surround, the black colours look lighter while the light colours remain relatively constant. As the surround luminance increases, the blacks begin to look darker, causing overall image contrast to increase.
36
11. Colour Order Systems Colour order systems arrange colours in a “space” – one that exists only in our imaginations – within which colours change continuously. Most of these systems arrange colours shading from dark at the bottom to light at the top, with hues arranged circumferentially and saturation increasing outward from a central achromatic axis. At the vertical extremes, with white above and black below, no saturation variation is possible. The maximum chromatic variation occurs at intermediate lightness levels. The outermost shell of the space resembles two lopsided cones joined at their bases with apices opposites.
11.1 The Munsell System The American colour teacher Albert H. Munsell (1858-1918) developed the Munsell Colour System in 1905 [5]. He wanted to create a system in which the spacing between each colour and its neighbour could be perceived as equal, i.e. a perceptually uniform system.
Figure 11.1. Munsell colour system (Images from www.adobe.com).
There are ten basic hues in the system. Five primary colours: red (R), yellow (Y), green (G), blue (B) and purple (P). And five intermediate colours: yellow-red (YR), green-yellow (GY), blue-green (BG), purple-blue (PB), and red-purple (RP) placed in between. Each of these ten hues are further subdivided by four decimal numbers: 2.5, 5, 7.5 and10, giving 40 hues in total. These hues are arranged in a circle around a central vertical neutral grey-value (N) axis where all have equal distances and are selected in a way that opposing pairs result in an achromatic mixture, see Figure 11.1 (left). Each colour is characterised by three attributes: Munsell Hue (described above), Munsell Value (N) and Munsell Chroma (C). The Munsell Value indicates the index of brightness in terms of a neutral grey scale and ranges from 0N for pure black to 10N for pure white. The Munsell Chroma is the gradation of saturation. The scale starts at 0 for neutral, but there is no arbitrary end to the scale. Maximum chroma can be somewhere in between 10 and 26 depending on the hue. Thus,
37
different hues have different number of chromatic steps. That is why the shape of the colour space is asymmetric. The notation for a colour in the Munsell System is written H u e Value/Chroma. For example 7.5YR 7/12 (an orange colour), where 7.5YR is the Munsell Hue, 7 is the Munsell Value and 12 is the Munsell Chroma. The colours are arranged in a colour atlas, The Munsell Book of Colour, from 1929. This edition is still in use today and contains 1200-1500 colour chips.
11.2 NCS The Swedish Natural Colour System (NCS) was introduced in 1979 by a team led by Anders Hård. The objective with NCS was to establish a colour system with which a user with normal colour vision could determine colours without the need for colour measuring instruments or colour samples. The NCS system is designed as an aid to defining, for example, the colour of a wall in a room purely on the basis of its perception. The system is based on six elementary colours: white (W), black (S), yellow (Y), red (R), blue (B) and green (G). And all other colours are then described in terms of these. The system possesses the external shape of a double-cone where Y, R, B and G occupy the circular base with evenly spaced positions. The tips of the double-cone are W (above) or S (below). In this three-dimensional model, called the NCS colour space, all imaginable surface colours can be placed. The double cone is also divided into two two-dimensional models, the NCS colour circle (a horizontal section through the colour space) and the NCS colour triangle (a vertical section through the colour space), see Figure 11.2.
Figure 11.2. The NCS colour circle and the NCS colour triangle (www.ncs.se).
The colours in the system are characterized by three attributes: NCS Colour Hue (H), NCS Blackness (S) and NCS Chromaticity (C). The NCS Colour Hues are defined on the basis of the basic colours yellow, red, blue and green shown in the colour circle. Each of the quadrants in the circle is further subdivided between two basic colours by a scale that expresses the portion of each colour as a percentage. For example, Y40R implies a yellow with 40% red, and B20 G implies a blue with 20% green. This allocation is based on the principle of similarity, that each colour is similar to a maximum of two chromatic elementary colours (in addition
38
to white and black) and that such a match can be quantitatively assessed down to an accuracy of 5%. The NCS Blackness indicates the proportion of black and the scale ranges from 0 (white) to 100 (black). And the NCS Chromaticity indicates the degree of chromaticity and also varies from 0 (achromatic colour) to 100 (full chromatic colour). This is shown in the colour triangle where all colours, which lie on the vertical lines, contain equal chromatic proportions. In the same way, all colours in the rows running parallel to the line between white and the observed colour contain equal proportions of black. NCS colour notations are based on how much a given colour seems to resemble the six elementary colours. In the NCS notation S 2030-Y90R, for example, 2030 indicates the nuance, i.e. the degree of resemblance to black (S) and to the maximum chromaticness (C); in this case, 20% blackness and 30% chromaticness. The hue Y90R indicates the portion of each colour as a percentage; in this case a yellow (Y) with 90% redness (90R). Purely grey colours lack colour hue and are only given nuance notations followed by -N as neutral. 0500-N is white and this is followed by 1000-N, 1500-N, 2000-N and so on to 9000-N which is black.
11.3 DIN The Deutsche Institut für Normung (DIN) system was developed in Germany by Manfred Richter and introduced in 1953. The objective was to create a colour system operating with the explicit variables of colour hue, saturation and brightness and as perceptively equidistant as possible. The DIN system has three variables: DIN Colour Hue (T), DIN Saturation (S) and DIN Darkness (D). They provide the coordinates for the three dimensional system that has the shape of a cone. The DIN Colour Hue is defined by means of a colour circle with 24 gradations. Hue varies from a value of T=1 (yellow) via red (7), blue (16), and green (22) to a green-yellow that has a value of T=24. Within the DIN colour-circle the DIN Saturation gradations commence with S = 6 and end at an achromatic point S = 0, and both colour-hue and saturation together form the colour type. The DIN Darkness is related to the luminous reflectance of the sample relative to an ideal sample (a sample that either reflects all or none if the incident energy at each wavelength) of the same chromaticity [1]. This enables the DIN system to associate colours not of the same brightness but of the same relative brightness. In terms of perception, this is more appropriate, since we tend to sense colours of differing colour-hue as being of equal value. The scale ranges from a value of 0 (white) to 10 (black). The notation for colours in the DIN system is written in the sequence T:S:D. For example 22.5:3.2:1.7 (a green colour), where 22.5 is the DIN Colour Hue, 3.2 is the DIN Saturation and 1.7 is the DIN Darkness. The colours are arranged in a colour atlas, the DIN Colour Chart 6164, which contains 600 colour samples (20 x28 mm).
39
11.4 OSA UCS The Optical Society of America Uniform Color Scales (OSA UCS) system was introduced in 1960. The aim in developing the OSA UCS colour order system was to determine a set of colour samples that, under appropriate viewing conditions, defined a perceptual uniform colour space. The OSA colour system has the form of a cubo-octohedron, which is the form resulting from slicing off all corners of a cube down to the midpoint of each edge yielding 12 corner points. The colours of the cubo-octahedron have been selected so that the distances between a colour sample and each of its 12 nearest neighbours are perceived as equally large colour differences. The position of a sample within this space is defined by the coordinates of three axes, which intersect each other at right angles: Lightness (L), Yellowness-Blueness (j) and Greenness-Redness (g). The j-axis does not exactly correspond to a yellow-blue axis. The reference j represents yellow at high lightness values. For negative values of j the axis separates blue from the violet region. Correspondingly, the positive values for g will not indicate green, instead this parameter separates the blue and green colours. And red does not lie at the end of the negative g scale, but pink. The OSA Lightness value is zero when the brightness corresponds to the background generally recommended for viewing the samples, it is positive when a colour is brighter than the background, and it is negative for a colour that is darker. The samples are arranged in an array along a vertical axis running from black to white, orthogonal to two chromatic axes, one of which runs roughly from red to green, the other from blue to yellow. In the 1978 report issued by the Committee for Uniform Color Scales, a total of 558 samples were colorimetrically specified, together with their exact coordinates. The objective of equal colour differences in all directions results in a very different type of colour order system. Perhaps due to its complex geometry, the OSA UCS is not very popular [1].
40
12. Terminology Here follows some important definitions of our perceptions of colour stimuli. A complete specification of a colour appearance requires five perceptual dimensions: brightness, lightness, colourfulness, chroma and hue.
12.1 Colour Definition of colour: Attribute of visual perception consisting of any combinations of chromatic and achromatic content. This attribute can be described by chromatic colour names such as yellow, orange, brown, red, pink, etc., or by achromatic colour names such as white, grey, black, etc., and qualified by bright, dim, light, dark, etc., or by combinations of such names. Note: Perceived colour depends on the spectral distribution of the colour stimulus, on the size, shape and structure, and surround of the stimulus area, on the state of adaption of the observer’s visual system and on the observer’s experience of the prevailing and similar situation of observations. There are eleven basic colour terms that can be subdivided into three categories: 1. achromatic colour terms (white, grey, black); and two varieties of chromatic colour terms 2. primary (red, yellow, green, blue) and 3. secondary (orange, purple, pink, brown) We can describe all the colours we can discriminate by using the chromatic primary colour terms red, yellow, green and blue, and their combinations.
12.2 Hue Definition of hue: Attribute of a visual sensation according to which an area appears to be similar to one of the perceived colours: red, yellow, green, and blue, or a combination of the two of them. Definition of achromatic colour: Perceived colour devoid of hue. Definition of chromatic colour: Perceived colour possessing a hue. Hue is often described with a ”hue circle”. In which one can find the unique hues; red, yellow, green and blue and their combinations.
12.3 Brightness and Lightness The attributes of brightness and lightness are very often interchanged, despite the fact that they have very different definitions. Definition of brightness: Attribute of a visual sensation according to which an area appears to emit more or less light.
41
Definition of lightness: The brightness of an area judged relative to the brightness of a similarly illuminated area that appears to be white or highly transmitting. Note: Only related colours exhibit lightness. The standard definition of lightness is given by the equation below:
Lightness =
Brightness Brightness (white)
Brightness refers to the absolute perception of the amount of light of a stimulus, while lightness can be thought of as the relative brightness. The visual system generally behaves as a lightness detector. Example: A newspaper when read indoors would have a certain brightness and lightness. When viewed side by side with standard office paper, the newspaper often looks slightly grey, while the office paper appears white. When the newspaper and office paper are brought outdoors on a sunny summer day, they would then have much higher brightness. Yet the newspaper still appears darker than the office paper as it has a lower lightness. The physical amount of light reflected from the newspaper might be more than a hundred times greater than the office paper was indoors, yet the relative amount of light reflected has not changed. Thus, the relative appearance between the two papers has not changed.
12.4 Colourfulness and Chroma Definition of colourfulness: Attribute of a visual sensation according to which the perceived colour of an area appears to be more or less chromatic. Note: For a colour stimulus of a given chromaticity and, in the case of related colours, of a given luminance factor, this attribute usually increases as the luminance is raised, except when the brightness is very high. Definition of chroma: Colourfullness of an area judged as a proportion of the brightness of a similarly illuminated area that appears white or highly transmitting. Note: For given viewing conditions and at luminance levels within the range of photopic vision, a colour stimulus perceived as a related colour, of a given chromaticity, and from a surface having a given luminance factor, exhibits approximately constant chroma for all levels of luminance except when the brightness is very high. In the same circumstances, at a given level of illuminance, if the luminance factor increases, the chroma usually increases. The standard definition of chroma is given by the equation below:
Chroma =
Colourfulness Brightness (white)
Colourfulness describes the amount or intensity of the hue of a colour stimulus and thus is an absolute perception. And chroma can be thought of as relative colourfulness just as lightness can be thought of as relative
42
brightness. The human visual system generally behaves as a chroma detector.
12.5 Saturation Definition of saturation: Colourfulness of an area judged in proportion to its brightness. Note: For given viewing conditions and at luminance levels within the range of photopic vision, a colour stimulus of a given chromaticity exhibits approximately constant saturation for all luminance levels, except when brightness is very high. The standard definition of saturation is given by the equation below:
Saturation =
Colourfulness Chroma = Brightness Lightness
Like chroma, saturation can be thought of as relative colourfulness. However, saturation is the colourfulness of a stimulus relative to its own brightness, while chroma is colourfulness relative to the brightness of a similarly illuminated area that appears white. For a stimulus to have chroma it must be judged in relation to other colours, while a stimulus seen completely in isolation can have saturation.
12.6 Related and Unrelated Colours The definition of colour is further enhanced with the notion of related and unrelated colours. Definition of related colour: Colour perceived to belong to an area of object seen in relation to other colours. Definition of unrelated colour: Colour perceived to belong to an area of object seen in isolation from other colours. The colours brown and grey only exists as related colours. It is impossible to find an isolated brown or grey stimulus, as evidenced by the lack of a brown or grey light source. These lights would appear either orange or white when viewed in isolation.
12.7 Achromatic and Chromatic Colours Definition of achromatic colours: When light reflection is flat across the spectrum, such as white, black or grey. Definition of chromatic colours: When some wavelengths are reflected more than others, as for example blue pigment.
43
13. References [1]
Fairchild, M. Colour Appearance Models, First Edition, AddisonWesley, Massachusetts (1998).
[2]
Field, G. Color and its Reproduction, Second edition, Sewickley GatfPress, (1999).
[3]
Giorgianni, E.J. and Madden, T.E. Digital Color Management – Encoding Solutions, Addison-Wesley, Massachusetts (1998).
[4]
Goldstein, E. B. Sensation and Perception, Sixth edition, Wadsworth publishing Company, Belmont, CA. (1998).
[5]
Kaiser, P.K. and Boynton, R.M. Human Colour Vision, Second Edition, Optical Society of America, Washington, D.C. (1996).
[6]
Wyszecki, G., and Stiles, W.S. Colour Science, Second Edition, John Wiley and Sons, New York (2000).
44