A set of colours aesthetically pleasant are described as harmonious in the language of human visual perception. As this notion encloses a subjective part, a psychophysical experiment was carried out to estimate the perception of colour harmony for combinations of paintings with the uniform colour of walls on which they are hung.The experiment, that involved 38 observers, was based on accurate colours built upon a specific colour flow. Participants were asked to judge the colour harmony of combinations of a sample of 7 selected paintings with backgrounds uniformly coloured in 3 different ranges of colours – achromatic colours, tones derived from the global average colour of the considered painting and tones derived from the complementary of the global average colour of the considered painting.Results demonstrate that the best colour harmony is obtained when the average colour of paintings is used to colour their background. The experiment presented in this paper clearly shows that the white colour usually used for walls in museums does not optimize the colour harmony.

In computer vision several methods are directly inspired by the human visual system. Visual attention is one of the main abilities a human eye uses to discover a new scene. This ability is based on the focus of attention principle, which enables to look at a particular point. The works presented in this paper describe how to simulate such an ability in computer vision. Color images are resampled according to this principle, in order to considerably decrease the amount of data to be processed. This resampling is done by using a concentric distribution of hexagonal cells instead of the rectangular cell grid generally provided by uniform sensors such as CCD cameras. Such a distribution is derived from the cone distribution on the retina and the result is encoded by using polar coordinates. In this way the information is more and more blurred in an isotropic way, when getting far and far from the focusing point. This resampling method then allows decreasing data number while it globally keeps all the information included in the image. In this way a new scene can be explored by focusing successively at a sequence of focusing points. This allows reproducing the human eye behavior that explores a new scene by saccades. Furthermore the resampled images can be used in order to set up image preprocessing. For example this resampling can be achieved before a preliminary step of segmentation. The resampled image is segmented in order to coarsely determine regions. Afterwards the segmentation step can be refined by combining the segmented resampled image and the original image. A comparison between results obtained in RGB and HSV spaces is also given on a set of images.

Understanding how colour is used by the human vision system is a widely studied research field. The field, though quite advanced, still faces important unanswered questions. One of them is the explanation of the unique hues and the assignment of color names. This problem addresses the fact of different perceptual status for different colors.Recently, Philipona and O'Regan have proposed a biological model that allows to extract the reflection properties of any surface independently of the lighting conditions. These invariant properties are the basis to compute a singularity index that predicts the asymmetries presented in unique hues and basic color categories psychophysical data, therefore is giving a further step in their explanation.In this paper we build on their formulation and propose a new singularity index. This new formulation equally accounts for the location of the 4 peaks of theWorld colour survey and has two main advantages. First, it is a simple elegant numerical measure (the Philipona measurement is a rather cumbersome formula). Second, we develop a colour-based explanation for the measure.

Pair comparison methods based on Case V of Thurstone's Law of Comparative Judgment are widely used to derive interval scales for perceptual image quality. A thorough treatment of the involved statistical errors is often neglected, even though this is the base for computing confidence intervals and other statistical tests. In this paper we show, that consequent error estimation through all steps of the data analysis provides a simple and reliable method to compute confidence intervals. Monte Carlo simulations are used to verify the results and to compare the proposed error estimation with other known methods.

The relationship between the spectral composition of light sources and the visual appearance of rendered scenes is a matter of practical relevance and assumes today particular significance with the advent of light sources of almost arbitrary spectral distribution, like modern LED based lighting. This relationship has only been studied for specific illuminants, like daylights, and systematic studies with other light sources are necessary. The aim of this work was to address this issue by studying, computationally, some chromatic effects of metamers of daylight illuminants. For each daylight with correlated color temperature (CCT) in the range 25 000 K – 4000 K a large set of metamers was generate using the Schmitt's elements approach. The metamers set was parameterized by the absolute spectral difference to the equi-energy illuminant E and by the number of non-zero spectral bands. The chromatic effects of the metamers were quantified by the CIE color rendering index CRI and by the CIELAB color gamut generated when rendering the Munsell set. It was found that although CRI decreases with, that is, as the illuminant spectrum becomes spectrally more structured, the largest values for the color gamut could be obtained only for large values of. Furthermore, the relationship between color gamut and number of non-zero bands showed that the largest gamuts were obtained with a small number of spectral bands. Thus, spectrally structured metamers produced low CRI but larger color gamuts, a result suggesting that appropriate spectral tuning may be explored in practical illumination when obtaining large chromatic diversity may be important.

The characterization of trichromatic cameras is usually done in terms of a device-independent color space, such as the CIE 1931 XYZ space. This is indeed convenient since it allows the testing of results against colorimetric measures. We have characterized our camera to represent human cone activation by mapping the camera sensor's (RGB) responses to human (LMS) through a polynomial transformation, which can be “customized” according to the types of scenes we want to represent. Here we present a method to test the accuracy of the camera measures and a study on how the choice of training reflectances for the polynomial may alter the results.

Common descriptors of light quality fail to predict the chromatic diversity produced by the same illuminant in different contexts such as images of natural scenes. The aim of this paper was to introduce a new index, capable of predicting illuminant-induced variations in the chromatic diversity off natural scenes. The spectral reflectance of each pixel of 50 images of natural scenes obtained using a hyperspectral imaging and the spectral reflectance of 1264 Munsell surfaces were converted into the CIELAB color space for each of the 55 illuminants and 5 light sources. The CIELAB volume was estimated by the convex hull method. The number of discernible colors was estimated by segmenting the CIELAB color volume into unitary cubes and by counting the number of non-empty cubes. High correlation was found between the CIELAB volume occupied by the Munsell surfaces, the number of discernible colors and CILEAB color volume of the colors of natural scenes. These results seem to indicate that a new illuminant chromatic diversity index based on natural scenes could be defined using the CIELAB volume of the Munsell surfaces.

In single-sensor digital imaging a color filter array, that is overlaid onto the image sensor, makes color images possible. Incident light rays become band-limited and each sensor element captures either red, green or blue light. Interpolating the missing two color components for each pixel location is known as demosaicing. This paper proposes to firstly derive an estimated luminance image by low-pass filtering the original mosaiced sensor image. In a second step a deconvolution technique re-sharpenes the blurred luminance approximation, so that it has the same spatial resolution as the original - but bandpassed - sensor image. Using the high-resolution luminance approximation the partial RGB colors from the mosaiced sensor image are transformed into a different color space that is more suitable for color interpolation. The new color space consists of least correlated color data, so that intra-channel interpolation errors have a reduced impact on inter-channel alignment, and therefore result into less prominent interpolation artifacts. Demosaicing is performed on the transformed color data separately for each plane, whereby again the luminance approximation, which encodes the aligned gradient direction of all color channels, regularizes the bilinear interpolation. Finally, the result is remapped into the RGB color space to obtain the demosaiced color image. Additionally, correlated multi-channel anisotropic diffusion is applied onto the demosaiced color image to further reduce interpolation artifacts and enable denoising. The proposed algorithm is evaluated and it is concluded that - although the image formation model could be verified - its performance heavily depends on the quality of the luminance approximation, i.e. the deconvolution method.

We propose a method for colorization of medical grayscale images using color learning. The colors are learned from a color image and predicted for a grayscale image. Earlier we introduced an efficient algorithm for image colorization which uses a dichromatic reflection model. The colorization algorithm is further developed in this study. First, we improve the algorithm performance by extending its capability to work with the grayscale images the contrast of which is lower than the contrast of the color images. Then, we propose a reliable technique to prevent negative contrast during colorization. In addition, we develop a simple approach for grayscale image colorization by a given RGB value. We give two medical applications of our algorithm: realistic color labeling of skin wounds and colorization of a dental cast models. In the former case we use grayscale images and labeling obtained after support vector classification as input data and for the latter application we use photometric stereo images.

Measuring the perceived quality of printed images is important to assess the performance of printers and to evaluate technology advancements. Image quality metrics have been proposed to objectively assess the quality of images, and new metrics are proposed continuously. However, since these metrics require digital inputs, applying these metrics to printed images are not straightforward. In order to accomplish this, the printed reproduction needs to be transformed into a digital copy.In this paper we propose a framework for applying image quality metrics to printed images, including the transformation to a digital format, image registration, and the application of image quality metrics. The proposed framework introduces less error and is significantly faster than another state of the art framework. Finally, the framework is used to evaluate a set of image quality metrics against subjective data.