Hans Brettel

Multispectral color image capture using a liquid crystal tunable filter

We describe the experimental setup of a multispectral color image acquisition system consisting of a professional monochrome CCD camera and a tunable filter in which the spectral transmittance can be controlled electronically. We perform a spectral characterization of the acquisition system taking into account the acquisition noise. To convert the camera output signals to device-independent color data, two main approaches are proposed and evaluated. One consists in applying regression methods to convert from the K camera outputs to a device-independent color space such as CIEXYZ or CIELAB. Another method is based on a spectral model of the acquisition system. By inverting the model using a principal eigenvector approach, we estimate the spectral reflectance of each pixel of the imaged surface.

Computerized simulation of color appearance for dichromats

We propose an algorithm that transforms a digitized color image so as to simulate for normal observers the appearance of the image for people who have dichromatic forms of color blindness. The dichromats color confusions are deduced from colorimetry, and the residual hues in the transformed image are derived from the reports of unilateral dichromats described in the literature. We represent color stimuli as vectors in a three-dimensional LMS space, and the simulation algorithm is expressed in terms of transformations of this space. The algorithm replaces each stimulus by its projection onto a reduced stimulus surface. This surface is defined by a neutral axis and by the LMS locations of those monochromatic stimuli that are perceived as the same hue by normal trichromats and a given type of dichromat. These monochromatic stimuli were a yellow of 575 nm and a blue of 475 nm for the protan and deutan simulations, and a red of 660 nm and a blue-green of 485 nm for the tritan simulation. The operation of the algorithm is demonstrated with a mosaic of square color patches. A protanope and a deuteranope accepted the match between the original and the appropriate image, confirming that the reduction is colorimetrically accurate. Although we can never be certain of anothers sensations, the simulation provides a means of quantifying and illustrating the residual color information available to dichromats in any digitized image.

Digital video colourmaps for checking the legibility of displays by dichromats

We propose replacement colourmaps that allow a designer to check the colours seen by protanopes and deu- teranopes. Construction of the colourmaps is based on the LMS specification of the primaries of a standard video monitor and has been carried out for 256 colours, including 216 colours that are common to many graphics applications of MS Windows and Macintosh computing environments.

Discrimination thresholds in the two-dimensional spatial frequency domain

In two experiments we have determined the discriminability between two sinusoidal gratings as a function of orientation and spatial frequency differences. Twelve orientation (15 degrees steps) and four spatial frequencies (2, 4, 8, 12 c/deg) were considered and corresponding discrimination thresholds were determined. Results indicated that: (a) spatial frequency discrimination thresholds did not significantly vary over all frequency and orientation positions, and averaged at +/- 1/8 octave limits. (b) Orientation thresholds only slightly increased with spatial frequency (from +/- 5 degrees at 2 c/deg to +/- 6 degrees at 12 c/deg). (c) A large and consistent anisotropy occurred with orientation thresholds showing most sensitivity in the horizontal and vertical directions, with largest thresholds at the oblique angles. No oblique effect was observed with spatial frequency thresholds. These results confirm past observations, providing an extensive set of threshold measurements in the two-dimensional spatial domain.

Spectral characterization of electronic cameras

In order to properly calibrate an electronic camera for a variety of illuminants it is necessary to estimate the spectral sensitivity of the camera. This spectral characterization is obtained by measuring a set of samples of known spectral reflectances and by inverting the resulting system of linear equations. In the presence of noise, this system inversion is not straightforward. We describe several approaches to this problem. In particular we show that the choice of samples is of great importance for the quality of the characterization, and we present an algorithm for the choice of a reduced number of samples.

Spatial tuning properties in human amblyopia cannot explain the loss of optotype acuity.

Strabismic amblyopes may exhibit in their squinting eye an almost normal contrast sensitivity to gratings although their optotype acuity is strongly impaired. This discrepancy cannot be accounted for by reduced selectivity of spatial-frequency channels or line-sensitivity functions. Alternative explanations for the loss of letter visual acuity as abnormal interactions between psycho-physical detector-mechanisms or defects in spatial phase sensitivity are discussed.

Multispectral image capture using a tunable filter

In this article we describe the experimental setup of a multispectral image acquisition system consisting of a professional monochrome CCD camera and a tunable filter in which the spectral transmittance can be controlled electronically. We have performed a spectral characterization of the acquisition system taking into account the acquisition noise. To convert the camera output signals to device-independent data, two main approaches are proposed and evaluated. One consists in applying regression methods to convert from the K camera outputs to a device- independent color space such as CIEXYZ or CIELAB. Another method is based on a spectral model of the acquisition system. By inverting the model using a Principal Eigenvector approach, we estimate the spectral reflectance of each pixel of the imaged surface.

Rating of tinted ophthalmic lenses

This study aims at establishing the relationship between the spectral transmission characteristics of tinted ophthalmic lenses and the visual percept of the observer looking through. We hypothesize that the rating of tinted lenses originates from what the observer sees rather than what the glass looks like. We develop a model for color acceptance by calculating the color distortions for a collection of colored surfaces that are representative of the real environment using the spectral reflectance of the surfaces and the spectral transmittance of the lens. Then we apply the model to a collection of real tinted ophthalmic lenses and derive statistical indices that describes the global color distortion induced by each lens. Preliminary results show a significant agreement between the subjective rating and the objective color distortion index. We conclude that our model allows to predict the acceptance/rejection judgement of observers for a tinted ophthalmic lens of known spectral transmittance.

Color display for dichromats

We have proposed colormaps to replace a widely used 216 colors palette in order to allow a designer with normal color vision to simulate the colors seen by dichromats. As dichromats lack one class of cone photo pigment, they confuse colors that differ only in the excitation of the missing class of photo pigment. The method is based on the LMS colorimetric system, which specifies colors in terms of the relative excitations of the cones. We have constructed a rule to reduce any set of confused colors to a single three- component specification. We have introduced a modification, assuming that the video display primaries and nominal white are representative of recent standards for Cathode Ray Tube monitors and that its video-transfer function is a power function with an exponent of 2.2. For everyday practice, replacing a normal palette by a reduced palette provides an immediate warning of possible losses of readability of a display by color-deficients.

Enhancement of Improperly Exposed Photographic Color Negatives

This paper addresses digital techniques used to automatically correct those color photographs whose range of transmittance densities is too large for visually acceptable image reproduction. The first step consists in calibrating the image acquisition devices, and results are shown for two different models. Then, we present a method inspired by photographic techniques that uses modified binary masks to enhance negatives of too high contrast. This method can be applied in an industrial environment such as in photographic mini-laboratories.