Colour vision: Is colour constancy real?

Optimisation of Convolution-Based Image Lightness Processing

In the convolutional retinex approach to image lightness processing, an image is filtered by a centre/surround operator that is designed to mitigate the effects of shading (illumination gradients), which in turn compresses the dynamic range. Typically, the parameters that define the shape and extent of the filter are tuned to provide visually pleasing results, and a mapping function such as a logarithm is included for further image enhancement. In contrast, a statistical approach to convolutional retinex has recently been introduced, which is based upon known or estimated autocorrelation statistics of the image albedo and shading components. By introducing models for the autocorrelation matrices and solving a linear regression, the optimal filter is obtained in closed form. Unlike existing methods, the aim is simply to objectively mitigate shading, and so image enhancement components such as a logarithmic mapping function are not included. Here, the full mathematical details of the method are provided, along with implementation details. Significantly, it is shown that the shapes of the autocorrelation matrices directly impact the shape of the optimal filter. To investigate the performance of the method, we address the problem of shading removal from text documents. Further experiments on a challenging image dataset validate the method.

A true color palette: binary metastable photonic pigments

Different from the traditional concept that binary photonic crystals can only reproduce mixed colors due to the simple superposition of the photonic band gaps, precisely addressable "true colors" obtained from volume fraction deviation of binary photonic crystals with metastable structures are reported here. Inspired by the mussels' adhesion and longhorn beetles' photonic scales, a binary metastable amorphous photonic crystal was obtained by enhancing the driving forces and customizing the surface roughness of building blocks to regulate the thermodynamic and dynamic factors simultaneously. By controlling the volume fraction of two building blocks, the tunable photonic bandgap varies linearly in the visible region. Furthermore, the "true violet" that cannot be obtained by conventional color mixing is reproduced with the particular ultraviolet characteristics of red photonic pigment's metastable structures, which complement the palette effect of "true colors". Meanwhile, due to the self-adhesion and post-modification of building blocks, the stability of photonic pigments is further improved. The binary photonic pigments not only solve the dilemma of mixed colors, but also realize the tunability and multiplicity of "true colors", offering a new choice for the color palette of the world.

Disentangling simultaneous changes of surface and illumination

Retinally incident light is an ambiguous product of spectral distributions of light in the environment and their interactions with reflecting, absorbing, and transmitting materials. An ideal color constant observer would unravel these confounded sources of information and account for changes in each factor. Scene statistics have been proposed as a way to compensate for changes in the illumination, but few theories consider changes of 3-dimensional surfaces. Here, we investigated the visual system's capacity to deal with simultaneous changes in illumination and surfaces. Spheres were imaged with a hyperspectral camera in a white box and their colors, as well as that of the illumination were varied along "red-green" and "blue-yellow" axes. Both the original hyperspectral images and replica scenes rendered with Mitsuba were used as stimuli, including rendered scenes with Glavens (Acta Psychologica, 2009, 132, 259-266). Observers viewed sequential, random pairs of our images, with either the whole scene, only the object, or only a part of the background being present. They judged how much the illuminant and object color changed on a scale of 0-100%. Observers could extract simultaneous illumination and reflectance changes when provided with a view of the whole scene, but global scene statistics did not fully account for their behavior, while local scene statistics improved the situation. There was no effect of color axis, shape, or simulated vs. original hyperspectral images. Observers appear to be making use of various sources of local information to complete the task.

Study of chromatic adaptation via neutral white matches on different viewing media

Two experiments were carried out to study the neutral white and the chromatic adaptation in human vision and color science. After matching neutral whites under different illuminants using both surface and self-luminous colors, the result were used to verify the previous study about the chromatic adaptation. Not all the white illuminants were found neutral even the adaptation time is long. The baseline illuminant of the two-step chromatic adaptation transform was found as the illuminant with the same chromaticity of the neutral white under it and depended on viewing medium in the present study. The results were also used as corresponding colors to derive models of the effective degree of chromatic adaptation, which were found highly associated with the chromaticity of the adapting illuminant.

Cortical Color and the Cognitive Sciences

Back when researchers thought about the various forms that color vision could take, the focus was primarily on the retinal mechanisms. Since that time, research on human color vision has shifted from an interest in retinal mechanisms to cortical color processing. This has allowed color research to provide insight into questions that are not limited to early vision but extend to cognition. Direct cortical connections from higher-level areas to lower-level areas have been found throughout the brain. One of the classic questions in cognitive science is whether perception is influenced, and if so to what extent, by cognition and whether a clear distinction can be drawn between perception and cognition. Since perception is seen as providing justification for our beliefs about properties in the external world, these questions also have metaphysical and epistemological significance. The aim of this paper is to highlight some of the areas where research on color perception can shed new light on questions in the cognitive sciences. A further aim of the paper is to raise some questions about color research that are in dire need of further reflection and investigation.

Perceptual Pragmatism and the Naturalized Ontology of Color

This paper considers whether there can be any such thing as a naturalized metaphysics of color-any distillation of the commitments of perceptual science with regard to color ontology. I first make some observations about the kinds of philosophical commitments that sometimes bubble to the surface in the psychology and neuroscience of color. Unsurprisingly, because of the range of opinions expressed, an ontology of color cannot simply be read off from scientists' definitions and theoretical statements. I next consider two alternative routes. First, conceptual pluralism inspired by Mark Wilson's analysis of scientific representation. I argue that these findings leave the prospects for a naturalized color ontology rather dim. Second, I outline a naturalized epistemology of perception. I ask how the correctness and informativeness of perceptual states is understood by contemporary perceptual science. I argue that the detectionist ideal of correspondence should be replaced by the pragmatic ideal of usefulness. I argue that this result has significant implications for the metaphysics of color.

Nest covering in plovers: How modifying the visual environment influences egg camouflage

Camouflage is one of the most widespread antipredator defences, and its mechanistic basis has attracted considerable interest in recent years. The effectiveness of camouflage depends on the interaction between an animal's appearance and its background. Concealment can therefore be improved by changes to an animal's own appearance, by behaviorally selecting an optimal background, or by modifying the background to better match the animal's own appearance. Research to date has largely focussed on the first of these mechanisms, whereas there has been little work on the second and almost none on the third. Even though a number of animal species may potentially modify their environment to improve individual-specific camouflage, this has rarely if ever been quantitatively investigated, or its adaptive value tested. Kittlitz's plovers (Charadrius pecuarius) use material (stones and vegetation) to cover their nests when predators approach, providing concealment that is independent of the inflexible appearance of the adult or eggs, and that can be adjusted to suit the local surrounding background. We used digital imaging and predator vision modeling to investigate the camouflage properties of covered nests, and whether their camouflage affected their survival. The plovers' nest-covering materials were consistent with a trade-off between selecting materials that matched the color of the eggs, while resulting in poorer nest pattern and contrast matching to the nest surroundings. Alternatively, the systematic use of materials with high-contrast and small-pattern grain sizes could reflect a deliberate disruptive coloration strategy, whereby high-contrast material breaks up the telltale outline of the clutch. No camouflage variables predicted nest survival. Our study highlights the potential for camouflage to be enhanced by background modification. This provides a flexible system for modifying an animal's conspicuousness, to which the main limitation may be the available materials rather than the animal's appearance.

Entropy, color, and color rendering

The Shannon entropy [Bell Syst. Tech J.27, 379 (1948)] of spectral distributions is applied to the problem of color rendering. With this novel approach, calculations for visual white entropy, spectral entropy, and color rendering are proposed, indices that are unreliant on the subjectivity inherent in reference spectra and color samples. The indices are tested against real lamp spectra, showing a simple and robust system for color rendering assessment. The discussion considers potential roles for white entropy in several areas of color theory and psychophysics and nonextensive entropy generalizations of the entropy indices in mathematical color spaces.

Color constancy by category correlation

Finding color representations that are stable to illuminant changes is still an open problem in computer vision. Until now, most approaches have been based on physical constraints or statistical assumptions derived from the scene, whereas very little attention has been paid to the effects that selected illuminants have on the final color image representation. The novelty of this paper is to propose perceptual constraints that are computed on the corrected images. We define the category hypothesis, which weights the set of feasible illuminants according to their ability to map the corrected image onto specific colors. Here, we choose these colors as the universal color categories related to basic linguistic terms, which have been psychophysically measured. These color categories encode natural color statistics, and their relevance across different cultures is indicated by the fact that they have received a common color name. From this category hypothesis, we propose a fast implementation that allows the sampling of a large set of illuminants. Experiments prove that our method rivals current state-of-art performance without the need for training algorithmic parameters. Additionally, the method can be used as a framework to insert top-down information from other sources, thus opening further research directions in solving for color constancy.

Generalization of color-difference formulas for any illuminant and any observer by assuming perfect color constancy in a color-vision model based on the OSA-UCS system

The most widely used color-difference formulas are based on color-difference data obtained under D65 illumination or similar and for a 10° visual field; i.e., these formulas hold true for the CIE 1964 observer adapted to D65 illuminant. This work considers the psychometric color-vision model based on the Optical Society of America-Uniform Color Scales (OSA-UCS) system previously published by the first author [J. Opt. Soc. Am. A 21, 677 (2004); Color Res. Appl. 30, 31 (2005)] with the additional hypothesis that complete illuminant adaptation with perfect color constancy exists in the visual evaluation of color differences. In this way a computational procedure is defined for color conversion between different illuminant adaptations, which is an alternative to the current chromatic adaptation transforms. This color conversion allows the passage between different observers, e.g., CIE 1964 and CIE 1931. An application of this color conversion is here made in the color-difference evaluation for any observer and in any illuminant adaptation: these transformations convert tristimulus values related to any observer and illuminant adaptation to those related to the observer and illuminant adaptation of the definition of the color-difference formulas, i.e., to the CIE 1964 observer adapted to the D65 illuminant, and then the known color-difference formulas can be applied. The adaptations to the illuminants A, C, F11, D50, Planckian and daylight at any color temperature and for CIE 1931 and CIE 1964 observers are considered as examples, and all the corresponding transformations are given for practical use.

Do people match surface reflectance fundamentally differently than they match emitted light?

We compared matches between colours that were both presented on a computer monitor or both as pieces of paper, with matching the colour of a piece of paper with a colour presented on a computer monitor and vice versa. Performance was specifically poor when setting an image on a computer monitor to match the colour of a piece of paper. This cannot be due to any of the individual judgments because subjects readily selected a matching piece of paper to match another piece of paper and set the image on the monitor to match another image on a monitor. We propose that matching the light reaching the eye and matching surface reflectance are fundamentally different judgments and that subjects can sometimes but not always choose which to match.

Representation or Context as a Cognitive Strategy in Colour Constancy?

If an identification task with colour constancy as its objective is carried out under drastically changing illumination, do people rely mainly on colour information or do they rely on other sources of information? This question suggested two hypotheses for testing: (i) context hypothesis: people rely mainly on colour information (spectral reflectance or illumination chromaticity) to achieve colour constancy; (ii) representation hypothesis: people rely mainly on all other clues associated with colour to achieve colour constancy, including form information (any shape elements) and space information (spatial coordinates or spatial correlation). Experiment 1 showed that form information was readily associated with colour information to produce implicit representation. This gave the best colour-constancy performance (95.72%) and the fastest processing speed, so it probably used a top–down process. However, it was also prone to error owing to assumptions. Space information was not readily associated with colour information so colour-constancy performance was halved (48.73%) and processing time doubled. When the subject was deprived of both information sources and only given colour information, this resulted in the longest reaction times and the worst colour-constancy performance (41.38%). These results clearly support the representation hypothesis rather than the context hypothesis. When all three clues were available at the same time, the order of preference was: symbol, location, then colour. Experiment 2 showed that when form information was the main clue, colour-constancy performance was conceptually driven and processed more quickly; this supports the representation hypothesis. However, when form information was not used, colour constancy was data-driven, processed more slowly, and achieved an inferior identification rate overall; this supports the context hypothesis.

Color rendering properties of interior lighting influenced by a switchable window

Switchable windows are being installed into modern buildings. We report on how the variable transmittance of gasochromic and electrochromic switchable windows changes the color rendering properties of the daylight passing through these windows. We present a series of color rendering examples. Correlated color temperature and color rendering indices are insufficient to describe the color properties of this filtered light, as these indices are beyond the applicability limits. We obtain more reliable results using the color shifts (deltaE) of different objects. We find that as the transmittance of the window changes, each surface color moves along a path of the CIELAB space in the same direction for both switchable units. The direction and distance moved differs among test-color samples. The appearance of an array of colored objects is strongly distorted at higher coloration states of the windows.

Colour constancy in diurnal and nocturnal hawkmoths

Diurnal and nocturnal hawkmoths have been shown to use colour vision for flower discrimination. Here, we present evidence that the nocturnal hawkmoth Deilephila elpenor and the diurnal hawkmoth Macroglossum stellatarum also have colour constancy. Colour constancy was shown in D. elpenor in two multiple-choice experiments with five different bluish colour patches under white and blue illumination and with five yellowish colour patches under white, blue and yellow illumination. The mechanism underlying colour constancy in both species was investigated in two dual-choice experiments. The choice behaviour is consistent with the use of the von Kries coefficient law. Although the moths have colour constancy, they react to the colour of the illumination. They make fewer choices when tested under the changed illumination, where they never receive a reward, compared with the training illumination. Even if colour constancy can be explained by a von Kries adaptation mechanism, the fact that the animals discriminate between different illuminations indicates that some additional process must be involved.

Separate neural definitions of visual consciousness and visual attention; a case for phenomenal awareness

What is the relation between visual attention and visual awareness? It is difficult to imagine being aware of something without attending to it, and by some, visual consciousness is simply equated to what is in the focus of attention. However, findings from psychological as well as from neurophysiological experiments argue strongly against equating attention and visual consciousness. From these experiments clearly separate neural definitions of visual attention and visual consciousness emerge. In the model proposed here, visual attention is defined as a convolution of sensori-motor processing with memory. Consciousness, however, is generated by recurrent activity between cortical areas. The extent to which these recurrent interactions involve areas in executive or mnemonic space depends on attention and determines whether a conscious report is possible about the sensory experience, not whether the sensory experience is there. This way, a strong case can be made for a pure non-cognitive form of seeing, independent of attentional selection, called phenomenal awareness. This can be dissociated from the reportable form, depending on attention, called access awareness. The hypothesis explains why attention and consciousness seem so intricately related, even though they are fully separate phenomena.

Painting with light by Rob and Nick Carter: dramatic failures of colour constancy in articulated scenes

Recent work by British artists Rob and Nick Carter uses kinetic lights illuminating abstract photographic prints to induce dramatic failures of colour constancy.

Surface-illuminant ambiguity and color constancy: effects of scene complexity and depth cues

Two experiments were conducted to study how scene complexity and cues to depth affect human color constancy. Specifically, two levels of scene complexity were compared. The low-complexity scene contained two walls with the same surface reflectance and a test patch which provided no information about the illuminant. In addition to the surfaces visible in the low-complexity scene, the high-complexity scene contained two rectangular solid objects and 24 paper samples with diverse surface reflectances. Observers viewed illuminated objects in an experimental chamber and adjusted the test patch until it appeared achromatic. Achromatic settings made tinder two different illuminants were used to compute an index that quantified the degree of constancy. Two experiments were conducted: one in which observers viewed the stimuli directly, and one in which they viewed the scenes through an optical system that reduced cues to depth. In each experiment, constancy was assessed for two conditions. In the valid-cue condition, many cues provided valid information about the illuminant change. In the invalid-cue condition, some image cues provided invalid information. Four broad conclusions are drawn from the data: (a) constancy is generally better in the valid-cue condition than in the invalid-cue condition: (b) for the stimulus configuration used, increasing image complexity has little effect in the valid-cue condition but leads to increased constancy in the invalid-cue condition; (c) for the stimulus configuration used, reducing cues to depth has little effect for either constancy condition: and (d) there is moderate individual variation in the degree of constancy exhibited, particularly in the degree to which the complexity manipulation affects performance.

Detecting changes of spatial cone-excitation ratios in dichoptic viewing

Spatial ratios of cone excitations produced by light reflected by different surfaces in a scene may provide the cue for discriminating changes in illuminant from changes in surface reflectances. To test whether these ratios can be computed across the two eyes, observers were presented with simulations on a computer-controlled monitor of pairs of juxtaposed or separated Munsell surfaces undergoing an illuminant change with a small change in cone-excitation ratios or a change with constant cone-excitation ratios. Surfaces were viewed either binocularly or dichoptically. Observers reliably discriminated the two changes in both viewing conditions, although less well dichoptically. Cone-excitation ratios, which may in principle be computed retinally, may also be computed cortically.

Spatial structure of cone inputs to color cells in alert macaque primary visual cortex (V-1)

The spatial structure of color cell receptive fields is controversial. Here, spots of light that selectively modulate one class of cones (L, M, or S, or loosely red, green, or blue) were flashed in and around the receptive fields of V-1 color cells to map the spatial structure of the cone inputs. The maps generated using these cone-isolating stimuli and an eye-position-corrected reverse correlation technique produced four findings. First, the receptive fields were Double-Opponent, an organization of spatial and chromatic opponency critical for color constancy and color contrast. Optimally stimulating both center and surround subregions with adjacent red and green spots excited the cells more than stimulating a single subregion. Second, red-green cells responded in a luminance-invariant way. For example, red-on-center cells were excited equally by a stimulus that increased L-cone activity (appearing bright red) and by a stimulus that decreased M-cone activity (appearing dark red). This implies that the opponency between L and M is balanced and argues that these cells are encoding a single chromatic axis. Third, most color cells responded to stimuli of all orientations and had circularly symmetric receptive fields. Some cells, however, showed a coarse orientation preference. This was reflected in the receptive fields as oriented Double-Opponent subregions. Fourth, red-green cells often responded to S-cone stimuli. Responses to M- and S-cone stimuli usually aligned, suggesting that these cells might be red-cyan. In summary, red-green (or red-cyan) cells, along with blue-yellow and black-white cells, establish three chromatic axes that are sufficient to describe all of color space.