Eva M. Valero

Multispectral synthesis of daylight using a commercial digital CCD camera

Performance of multispectral devices in recovering spectral data has been intensively investigated in some applications, as in spectral characterization of art paintings, but has received little attention in the context of spectral characterization of natural illumination. This study investigated the quality of the spectral estimation of daylight-type illuminants using a commercial digital CCD camera and a set of broadband colored filters. Several recovery algorithms that did not need information about spectral sensitivities of the camera sensors nor eigenvectors to describe the spectra were tested. Tests were carried out both with virtual data, using simulated camera responses, and real data obtained from real measurements. It was found that it is possible to recover daylight spectra with high spectral and colorimetric accuracy with a reduced number of three to nine spectral bands.

Spectral-daylight recovery by use of only a few sensors.

Linear models have already been proved accurate enough to recover spectral functions. We have resorted to such linear models to recover spectral daylight with the response of no more than a few real sensors. We performed an exhaustive search to obtain the best set of Gaussian sensors with a combination of optimum spectral position and bandwidth. We also examined to what extent the accuracy of daylight estimation depends on the number of sensors and their spectral properties. A set of 2600 daylight spectra [J. Opt. Soc. Am. A 18, 1325 (2001)] were used to determine the basis functions in the linear model and also to evaluate the accuracy of the search. The estimated spectra are compared with the original ones for different spectral daylight and skylight sets of data within the visible spectrum. Spectral similarity, colorimetric differences, and integrated spectral irradiance errors were all taken into account. We compare our best results with those obtained by using a commercial CCD, revealing the CCDs potential as a daylight-estimation device.

Recovering fluorescent spectra with an RGB digital camera and color filters using different matrix factorizations

The aim of a multispectral system is to recover a spectral function at each image pixel, but when a scene is digitally imaged under a light of unknown spectral power distribution (SPD), the image pixels give incomplete information about the spectral reflectances of objects in the scene. We have analyzed how accurately the spectra of artificial fluorescent light sources can be recovered with a digital CCD camera. The red-green-blue (RGB) sensor outputs are modified by the use of successive cutoff color filters. Four algorithms for simplifying the spectra datasets are used: nonnegative matrix factorization (NMF), independent component analysis (ICA), a direct pseudoinverse method, and principal component analysis (PCA). The algorithms are tested using both simulated data and data from a real RGB digital camera. The methods are compared in terms of the minimum rank of factorization and the number of sensors required to derive acceptable spectral and colorimetric SPD estimations; the PCA results are also given for the sake of comparison. The results show that all the algorithms surpass the PCA when a reduced number of sensors is used. The experimental results suggest a significant loss of quality when more than one color filter is used, which agrees with the previous results for reflectances. Nevertheless, an RGB digital camera with or without a prefilter is found to provide good spectral and colorimetric recovery of indoor fluorescent lighting and can be used for color correction without the need of a telespectroradiometer.

Color-signal filtering in the Fourier-frequency domain

We have analyzed the Fourier-frequency content of spectral power distributions deriving from three types of illuminants (daylight, incandescent, and fluorescent) and the color signals from both biochrome and nonbiochrome surfaces lit by these illuminants. As far as daylight and the incandescent illuminant are concerned, after filtering the signals through parabolic (low-pass) filters in the Fourier-frequency domain and then reconstructing them, we found that most of the spectral information was contained below 0.016 c/nm. When fluorescent illuminants were involved, we were unable to recover either the original illuminants or color signals to any satisfactory degree. We also used the spectral modulation sensitivity function, which is related to the human visual systems color discrimination thresholds, as a Fourier-frequency filter and obtained consistently less reliable results than with low-pass filtering. We provide comparative results for daylight signals recovered by three different methods. We found reconstructions based on linear models to be the most effective.

Combining transverse field detectors and color filter arrays to improve multispectral imaging systems

This work focuses on the improvement of a multispectral imaging sensor based on transverse field detectors (TFDs). We aimed to achieve a higher color and spectral accuracy in the estimation of spectral reflectances from sensor responses. Such an improvement was done by combining these recently developed silicon-based sensors with color filter arrays (CFAs). Consequently, we sacrificed the filter-less full spatial resolution property of TFDs to narrow down the spectrally broad sensitivities of these sensors. We designed and performed several experiments to test the influence of different design features on the estimation quality (type of sensor, tunability, interleaved polarization, use of CFAs, type of CFAs, number of shots), some of which are exclusive to TFDs. We compared systems that use a TFD with systems that use normal monochrome sensors, both combined with multispectral CFAs as well as common RGB filters present in commercial digital color cameras. Results showed that a system that combines TFDs and CFAs performs better than systems with the same type of multispectral CFA and other sensors, or even the same TFDs combined with different kinds of filters used in common imaging systems. We propose CFA+TFD-based systems with one or two shots, depending on the possibility of using longer capturing times or not. Improved TFD systems thus emerge as an interesting possibility for multispectral acquisition, which overcomes the limited accuracy found in previous studies.

Unsupervised illuminant estimation from natural scenes: an RGB digital camera suffices

A linear pseudo-inverse method for unsupervised illuminant recovery from natural scenes is presented. The algorithm, which uses a digital RGB camera, selects the naturally occurring bright areas (not necessarily the white ones) in natural images and converts the RGB digital counts directly into the spectral power distribution of the illuminants using a learning-based spectral procedure. Computations show a good spectral and colorimetric performance when only three sensors (a three-band RGB camera) are used. These results go against previous findings concerning the recovery of spectral reflectances and radiances, which claimed that the greater the number of sensors, the better the spectral performance. Combining the device with the appropriate computations can yield spectral information about objects and illuminants simultaneously, avoiding the need for spectroradiometric measurements. The method works well and needs neither a white reference located in the natural scene nor direct measurements of the spectral power distribution of the light.

Evaluating logarithmic kernel for spectral reflectance estimation—effects on model parametrization, training set size, and number of sensor spectral channels

In this work, we evaluate the conditionally positive definite logarithmic kernel in kernel-based estimation of reflectance spectra. Reflectance spectra are estimated from responses of a 12-channel multispectral imaging system. We demonstrate the performance of the logarithmic kernel in comparison with the linear and Gaussian kernel using simulated and measured camera responses for the Pantone and HKS color charts. Especially, we focus on the estimation model evaluations in case the selection of model parameters is optimized using a cross-validation technique. In experiments, it was found that the Gaussian and logarithmic kernel outperformed the linear kernel in almost all evaluation cases (training set size, response channel number) for both sets. Furthermore, the spectral and color estimation accuracies of the Gaussian and logarithmic kernel were found to be similar in several evaluation cases for real and simulated responses. However, results suggest that for a relatively small training set size, the accuracy of the logarithmic kernel can be markedly lower when compared to the Gaussian kernel. Further it was found from our data that the parameter of the logarithmic kernel could be fixed, which simplified the use of this kernel when compared with the Gaussian kernel.

Spectral-reflectance linear models for optical color-pattern recognition

We propose a new method of color-pattern recognition by optical correlation that uses a linear description of spectral reflectance functions and the spectral power distribution of illuminants that contains few parameters. We report on a method of preprocessing color input scenes in which the spectral functions are derived from linear models based on principal-component analysis. This multichannel algorithm transforms the red-green-blue (RGB) components into a new set of components that permit a generalization of the matched filter operations that are usually applied in optical pattern recognition with more-stable results under changes in illumination in the source images. The correlation is made in the subspace spanned by the coefficients that describe all reflectances according to a suitable basis for linear representation. First we illustrate the method in a control experiment in which the scenes are captured under known conditions of illumination. The discrimination capability of the algorithm improves upon the conventional RGB multichannel decomposition used in optical correlators when scenes are captured under different illuminant conditions and is slightly better than color recognition based on uniform color spaces (e.g., the CIELab system). Then we test the coefficient method in situations in which the target is captured under a reference illuminant and the scene that contains the target under an unknown spectrally different illuminant. We show that the method prevents false alarms caused by changes in the illuminant and that only two coefficients suffice to discriminate polychromatic objects.

Detailed experimental characterization of reflectance spectra of Sasakia charonda butterfly using multispectral optical imaging

Abstract. A multispectral acquisition system to examine the bidirectional reflectance distribution function of structurally colored biological materials in the visible range is presented. We focus on the purple-blue and white-pearl wing scales of the male butterfly Sasakia charonda. Multispectral imaging was done by changing the illumination angular position around the sample as well as that of the specimen around the multispectral sensor axis. Reflectance spectra were transformed to color coordinates and visualized in different color spaces. Spectral analysis shows distinct iridescent patterns in purple-blue and white-pearl scales. Colorimetric analysis indicates that purple-blue scales enhance blue coloring and exhibit higher color saturation. Principal component analysis reveals that the number of principal components that account for more than 99% of reflectance variability was higher in white-pearl scales. This suggests a higher spectral complexity in their spatial color pattern formation. Reconstruction of reflectance spectra from the principal components is discussed. We conclude that multispectral imaging provides new insights into spatial reflectance mapping that result from the combination of structural colorations and variable amounts of absorption pigments.

Stochastic independence of color-vision mechanisms confirmed by a subthreshold summation paradigm

We have used a subthreshold summation protocol to analyze spatial color-color interaction. By means of a CRT color monitor, we measured the threshold contours for a spatial frequency of 0.5 cycles/degree. Heterochromatic flicker photometry was used to obtain isoluminance. The results suggest that the blue-yellow (b-y) and red-green (r-g) contrast thresholds remained unchanged by the addition of fixed r-g and b-y subthreshold pedestals. Our subthreshold summation data then support the stochastic independence of colorvision mechanisms derived from Mullen and Sankerallis work [Vision Res. 39, 733 (1999)] despite the differences that exist between the two experimental methods.