Showing papers presented at "Color Imaging Conference in 1995"

Multispectral Color System with an Encoding Format Compatible to the Conventional Tristimulus Model.

Abstract: Based on the discussion of the disadvantages of today’s color analysis in an open system architecture, a fundamentally improved reproduction system is proposed. This includes multispectral sampling and multispectral color information transport via the communication net using an encoded multispectral format compatible to conventional tristimulus color values.

Reconstructing Images from Trichromatic Samples: From Basic Research to Practical Applications.

Abstract: A pressure regulator for fluid systems that includes dual pressure ranges wherein a first pressure responsive controller controls pressure at a first preselected pressure value and wherein a second pressure responsive controller controls pressure at a second preselected pressure value responsive to a feedback signal to switch the pressure ranges.

Fast Separation of Reflection Components and its Application in 3D Shape Recovery.

Abstract: In this paper we present a fast and robust method to separate reflection components in color images Thus the method is also suitable for various industrial and scientific applications The technique operates with a twodimensional color representation Furthermore, we show that our approach improves the experimental results of a fast three-dimensional surface reconstruction scheme

Appearance Match between Soft Copy and Hard Copy under Mixed Chromatic Adaptation.

Abstract: Human visual system is partially adapted to the CRT monitor's white point and partially to the ambient light, when comparing a soft-copy image with the reproduced hard-copy image. The visual experiments were performed on the effect of the ambient lighting under mixed chromatic adaptation. It was found that human visual system is 40% to 60% adapted to CRT monitor's white point and the rest to the ambient light. This adaptation ratio itself was found to be independent of the luminance level of the ambient lighting for the portrait images. Practical method for appearance match between soft-copy and hard-copy is presented in this paper. This method is fundamentally based on a simple von Kries' adaptation model and in addition, takes into account of the human visual system's "partial adaptation."

The Effect of image content on color difference perceptibility

Abstract: Considerable work has been accomplished regarding the perceptibility of color differences for simple images. Less is known regarding color difference perceptibility when complex images are involved. In this research, psychophysical experimentation was conducted to determine if image content, in terms of memory color or image component size, is a significant factor in color difference perceptibility. A portrait, a nature scene, an image containing strictly manmade objects, and four mosaic images composed of different size patches were examined. The experimental results suggest that image component size affected the perceptibility of color differences while the presence of memory colors had no conclusive effect.

Linear Models and Computational Color Constancy.

Abstract: Linear models appear in many recent color constancy theories; however, they can play two quite different roles. They may either occur as a direct component of the computational strategy, as in the case of the Maloney-Wandell algorithm; or they may appear as necessary part of the theoretical development, but not as part of the computation perse, as in the case of algorithms based on spectral sharpening (Finlayson, Drew, Funt). This paper surveys recent work in color constancy and concludes that in many circumstances the second role may be more appropriate than the first.

Optimized Basis Functions for Coding Reflectance Spectra Minimizing the Visual Color Difference.

Abstract: The colors of objects depend on their reflectance spectra and on the spectrum of their illumination. To allow color reproduction for different light sources it is necessary to record and store more information on the reflectance spectrum than only the three-stimulus values supplied by usual cameras. A common method to compress the overwhelming amount of data to describe reflectance spectra is transform coding. The spectra are represented as linear combinations of an orthonormal basis vector set. Some work has already been carried out concerning the problem of choosing appropriate basis functions. Exact reproduction can be achieved for a limited number of N different light sources using 3N basis functions.1 If the number of light sources increases, small color errors must be admitted to maintain a practicable number of basis functions. The use of Fourier basis functions is a strategy that exploits the band limited characteristic of reflectance spectra. A different approach aims for a minimization of the mean squared error (MSE) between the original and the reconstructed spectrum using as few basis vectors as possible. Here an analytical solution exists called the principal component analysis (PCA) which depends on the statistical properties of the spectra set to be represented2,3. It offers the best data compression efficiency because of optimal energy concentration in few components. Nevertheless a correct solution would aim for a different error measure to minimize. The MSE between the spectra allows no statement concerning the visual color difference. It disregards the human visual system perceiving the colors as well as it does not take different illuminations into account. To analyze the visual color difference resulting from coding a reflectance spectrum it is necessary to calculate the tristimulus color values under all of the considered illuminants. These color triples must be transformed into a visually uniform color space before calculating the difference between the original and the coded reflectance spectrum. Transformation into standardized visually uniform color spaces implies nonlinear operations. The error measure might contain nonlinearities as well. Thus minimizing such an error measure one encounters a nonlinear quality criterion for which no analytical solution for the optimal basis function set can be given. Therefore an unconstrained stochastic optimization algorithm is applied to calculate sets of basis functions in order to minimize the visual color difference.

A Grid-Based Method for Predicting the Behaviour of Colour Printers.

Abstract: A new grid-based method is proposed for predicting the behaviour of colour printers. The method takes into account the varying density of dots as well as the light diffusion in the paper by defining at different intensity levels different colorimetric values for the printed ink as well as for the paper white. Since the model integrates both the varying density of partly overlapping ink dots and the light diffusion in the underlying substrate, the obtained predictions are more accurate than those obtained with surface based colour prediction methods described in the literature.

Color Reproduction Algorithms and Intent

Abstract: The effect of image type on systematic differences from strict colorimetric reproduction is investigated and discussed Different situations, called “color intents,” are described We have identified four basic color intents Color reproduction algorithms for each intent are presented, together with guidelines for selecting the appropriate intent

Aesthetic Considerations in Tone and Color Management.

Abstract: Color-management engineers have concentrated their attention on scientific issues (colorimetry, adaptation) in their efforts to provide consistent color reproduction on a variety of media and devices. Aesthetic issues have mostly been left to the user. However, some of these issues are related to the characteristics of media and are amenable to analysis and automation. A study of tone management exemplifies this approach.

Precision of Color Matches and Accuracy of Color-Matching Functions in Cross-Media Color Reproduction.

Abstract: A visual experiment was completed to evaluate the degree of observer metamerism in color matches between typical color reproduction media. The results illustrate that current CIE standard colorimetric observers provide reasonable estimates of average color matches and that the range of color mismatches encountered in cross media color reproduction due to observer variability can be very large, on the order of 10 CIELAB units.

Perceived Grey-Levels in Complex Configurations.

Abstract: We present a non-recursive integral-equation model that predicts perceived grey levels in complex achromatic displays as a function of the physical luminances of individual pixels in the display. The model incorporates spatially local luminance adaptation mechanisms, contrast gain controls and spatial distance dependent weights on lateral connections, and linear summation of the induced effects of individual surrounding pixels.

A Local Color Correlation Measure for Color Image Comparison.

Abstract: In image analysis we are faced to numerous problems more or less difficult to solve. The image comparison is one of the most interesting problems to solve because it involves two fields of study. The first one relates to image analysis in accordance with human visual perception inorder to reach results that correspond to the human visual judgement. The second field of study deals with image processing techniques such as segmentation, quantization or compression. As far as an objective image comparison is concerned, we can first of all evaluate the result of the process itself with respect to image quality, we can likewise compare the results of two processes working on the same image with respect to achieved images. This will enable us to understand or to evaluate the reasons why a particular process is not optimal as for image quality and to improve it therefore.1 Although image comparison processes have largely evolved during the last few years, they do not enough take into account visual parameters to be considered being relevant with reference to visual judgement.2 Even if the experience and theory show that it is extremely difficult to define an objective process that involves the most significant phenomena of the visual judgement, it has been shown that we can nevertheless define an heuristic process from some basic characteristics linked to human visual perception.3,4 In this article we propose to use a color correlation measure in order to define a subjective method for color image quality evaluation.5 Moreover, rather than defining a global measure of image differences such as the sum of squared differences, we propose first to compute this color correlation measure for each pixel taking into account its neighborhoods, secondly to display all these measures like a comparison image and thirdly to define a global measure according to the spatial distribution and to the amplitude of these local measures.6 As far as such a process is concerned, we can show that our measure is relevant in terms of human visual judgement. Furthermore we can put stress on image areas for which the correlation measure is the most important, A Local Color Correlation Measure for Color Image Comparison

Limitations in Communicating Color Appearance with the ICC Profile Format.

Abstract: The ICC format has been widely adopted as an industry standard for communicating color and a context has been presented for implementing and interpreting this standard. There are three major implications of the way the ICC profile format has been publicly presented for use in communicating color; user interface and application development complexity, possible quality limitations due to undefined transforms, and the possible quality degradation of business graphics. Some specific implications for peripheral vendors will also be discussed.

Using Reflectance Models for Surface Estimation.

Abstract: We suggest dichromatic reflection models adequate for describing surface-spectral reflectances of a variety of materials. First we describe the standard dichromatic reflection model for inhomogeneous dielectric materials. Next the generalized models are shown for describing the spectral reflectances of the materials for which the standard model is inadequate. The reflection characteristics are analyzed on a chromaticity diagram. We demonstrate practical use of the reflectance models in computer graphics and color constancy.

Chromatic Aberration, Linear Models, and Matching Color Images.

Abstract: near the accommodated wavelength can have detectable contrast in the retinal image, which implies that high spatial-frequency components play little role in color and contrast perception. Second, in the moderate spatial-frequency range, from 520 cpd, when the observer is accommodated to the yellow or green part of the spectrum, the visual system is dichromatic: there is no contrast in the short-wavelength receptor class. Perhaps most important, the OTF we calculated suggests an improved procedure for matching color images. The conventional method of setting point-by-point matches between images fails to account for the fact that image points on different displays may not have same pointspread function on the retina Since the spatial patterns on the retina from individual points on the displays do not match, one cannot match the retinal images of two points simply by adjusting the intensities of the three display primaries. Instead, to equate photo-pigment absorptions between images on different displays, one must adjust the primary intensities in corresponding spatial-frequency bands. (We describe this procedure in more detail below.) Because the OTF depends on the wavelength of the corneal image (as wed as its spatial frequency), using it to compute photoreceptor responses can be computationally quite expensive. When an image arises from a natural scene, representing the surface and illuminant spectral functions with finite-dimensional linear models greatly simplifies the computation. In that case, a simpler OTF can be computed that depends not on the wavelength of the corneal image but only on the weights of the basis functions that model the image. The number of weights win in most cases be much smaller than the number of wavelength samples, which is why the computation becomes so much less expensive. This simpler OTF can also be used to predict matches of color images on emissive displays. This is because emissive displays can be represented with a three-di-mensional linear model. We presented our algorithm for color matching on emissive displays in an earlier paper.7 Here we show that this use of the OTF is a special case of the OTF that arises from representing surface and illuminant functions with linear models. The remainder of the paper is organized as follows. In the next section, we introduce notation for the OTF and show how to use it to predict photoreceptor responses. Next, we introduce linear models for surface and illuminant spectral functions and show how their use simplifies the prediction of photoreceptor responses. Then we show how our Abstract

Image-designing and Color Management for Digital Still Camera and Printer System.

Abstract: We will describe tone-reproduction and color-reproduction of newly-developed digital still camera, DS-505, and Pictrography 3000 system. The followings are main content. (1) Comparison of Neugebauer’s Q-factors of digital still camera and typical negative film: The ascending order of Q-factor is 3-CCD camera, 1-CCD camera and typical negative film. (2) Aimed image-reproduction: The image, reproduced using professional-use negative and color paper, has been adopted as aimed image-reproduction. How to optimize various parameters to keep good image quality will be also described. (3) Color management system: To keep good image quality and data-exchangeability, EXIF (Exchangeable Image Data Format), proposed to ISO/TC42/ WG18, has been adopted. How to manage color quality parameters will be discussed.