In this paper we investigate the impact of colorimetric observer categories on the prediction of the average suprathreshold color difference perception. The observer categories were obtained from an observer classification experiment, while the color difference data were obtained from an experiment involving a liquid crystal display (LCD) with fluorescent backlight. The same observer panel with normal color vision participated in both experiments. Results obtained from the observer classification experiment were consistent with the average observer threshold for color difference judgment. This analysis demonstrates that the observer categories, determined based on individual differences in cone spectral sensitivities (and thus color matching functions), have an influence on the prediction of average suprathreshold color difference perception for a given observer population.

The paper proposes a novel approach to analyze the dichromatic color vision defects from a point of spectral responses based on the projection theory of spectral space to/from 2-D dichromatic Human Visual Sub-Space. The visible spectra to the dichromats (protanopes, deutanopes, and tritanopes) are extracted from an n-dimensional spectral input with the 2-D version of Matrix-R notated as R_dichro. Since the matrix R_dichro is an identical and invariant mapping operator inherent in human vision that is independent of any linear transformation or any illuminant, the fundamental spectra C*_dichro sensed with matrix R_dichro are also inherent in the dichromats. The lost spectra are easily obtained as a difference in the fundamentals between the normals and the the dichromats. These lost spectral profiles tell us why the color appearances are similar to the protanopes and deutanopes, and dissimilar to the tritanopes. The perceived colors are simulated based on the two hypotheses of substitution and nulling processes.

Professional photographers compose and process an image to emphasise the image's subject. Images with high salience, where a region is highly distinct from its background, are perceived to be of much greater quality in panel tests. Because of technical and expertise considerations, “average” camera users often capture images that have a lesser salience, thereby decreasing the image's appeal.The standard workflow to increase the perceived salience of an image's main subject consists in identifying the region of interest, and processing that region according to a set of rules. The level of analysis and processing can greatly vary, from increasing saturation or sharpness to identifying semantic concepts, e.g., faces, and employ a complex, tailored, modification.This is a delicate problem to approach: saliency prediction algorithms are currently not precise enough, and region classification is necessarily limited to a few specific classes. Furthermore, the variety of content often precludes the usage of a fixed set of rules in the enhancement step.Rather than attempting to predict saliency in images, we propose that important regions are somewhat distinct from their surroundings and can be identified by features that are spatially compact, in addition to standard compositional cues. Having identified the region of interest, we provide an enhanced image by increasing the values of its compact feature(s), i.e., increasing the perceived saliency of the region of interest. Preference studies indicate our modified images are significantly preferred to the original ones.

A paired comparison psychophysical experiment was conducted to investigate the perception of chromatic noise. Interestingly, chromatic noise on a grey patch was less visible than on chromatic patches. Among chromatic patches, chromatic noise on a purple patch was the most visible and chromatic noise on orange, yellow, or green patch was less visible. Then a heterochromatic brightness matching experiment was conducted and it was suggested that this perception of chromatic noise could be explained by the Helmholtz-Kohlrausch effect. The gradient of the luminance of the same brightness was shown to have a correlation with the chromatic noise visibility. Thus the chromatic noise was perceived not only as chromatic noise but also as brightness noise that should be more sensitive for the human vision. Due to the dependency of the Helmholtz-Kohlrausch effect on hue and chroma, the visibility of chromatic noise should depend on the colors of the patches.

An accurate image-difference measure would greatly simplify the optimization of imaging systems and image processing algorithms. The prediction performance of existing methods is limited because the visual mechanisms responsible for assessing image differences are not well understood. This applies especially to the cortical processing of complex visual stimuli.We propose a flexible image-difference framework that models these mechanisms using an empirical data-mining strategy. A pair of input images is first normalized to specific viewing conditions by an image appearance model. Various image-difference features (IDFs) are then extracted from the images. These features represent assumptions about visual mechanisms that are responsible for judging image differences. Several IDFs are combined in a blending step to optimize the correlation between image-difference predictions and corresponding human assessments.We tested our method on the Tampere Image Database 2008, where it showed good correlation with subjective judgments. Comparisons with other image-difference measures were also performed.

Paired comparison experiments are frequently used to gather observer preference data in many areas of image enhancement. However, due to the large quantity of comparisons each individual must complete, these experiments are typically carried out with few observers. Taking this method onto the web is a quick way of gaining a larger number of observers and preference judgements. This work examines the validity of web based paired comparisons and whether the loss of control over viewing conditions causes significantly different results.

We propose an automatic procedure for color scanner profile selection, which greatly improves the color quality of the final scanned images, and also the user experience. By just analyzing a small number of initial rows of the scanned image, our algorithm is capable of distinguishing among the available color profiles, and then it uses the one that better fits with the substrate of the original. The selection can be done “on the fly” with no time penalty for the user, and the color results obtained are improved when compared to a scanner workflow where no color profiles are used

Incomplete paired comparison is an important technique for color-imaging problems because it can avoid observers to compare every possible pairs since the number of paired comparisons for n stimuli is n(n-1)/2 which becomes prohibitive for large values of n. However, the experimental designer often struggles with questions such as what is the smallest limit the proportion of paired comparisons included that will still allow reliable estimations of scale values? Fortunately a Monte-Carlo computational simulation is carried out with a model of an ideal observer and the results shows that the proportion of paired comparisons that is included is more critical than the number of observers who make those observations [1]. This work aims to test the results from computational simulation with 25 real observers and 10 stimuli from the gray scale. The work suggests when each observer estimates the same proportion of paired comparisons included the more proportion of pairs and number of observers, the more accurate scale values will be produced and the proportion of pairs is more critical than the number of observers who make those observations, which quite agrees with the findings from the computational simulation. The work also suggests when the each observer estimates a different proportion of paired comparisons the more proportion of paired comparisons will not always produce a more accurate scale values.

In color-imaging science researchers frequently conduct experiments that are evaluated by measuring a number of ΔE values. Often the effect of various experimental parameters is evaluated by comparing the means of two or more sets of ΔE values. Frequently, a t-test is used but correct application of the t-test requires that a number of assumptions (such as the data being normally distributed or the sample size is big enough) are satisfied. In this article, an improved T statistic has been used in a case study where the data are skewed and do not have a normal distribution while at the same time the sample size is small. The importance of this assumption comes to mind when we realize that most of the derived data from the color-imaging science are not applicable for t-test because of not following a normal distribution.

A blueprint is a type of paper-based reproduction. For almost a century blueprint was the only low cost process available for copying drawings. Despite today this technology has been largely superseded by digital printing, it is still being used in emerging countries. When making copies of those documents, the colored background often makes difficult reading the information in the scanned plot. Hence, having an accurate way to transform the background to white enhances the look of the plot, plus saves ink when printing. This paper introduces an algorithm for removing the background color from a blueprint scanned image, comprising processing the image in a colorimetric space by determining the colorimetric values of the background color of the image, computing a linear chromatic adaptation transform for transforming the background-color colorimetric values to values of a target white point, and applying the computed chromatic adaptation transformation to the image.