Illuminant estimation in multispectral imaging

Abstract

With the advancement in sensor technology, the use of multispectral imaging is gaining wide popularity for computer vision applications. Multispectral imaging is used to achieve better discrimination between the radiance spectra, as compared to the color images. However, it is still sensitive to illumination changes. This study evaluates the potential evolution of illuminant estimation models from color to multispectral imaging. We first present a state of the art on computational color constancy and then extend a set of algorithms to use them in multispectral imaging. We investigate the influence of camera spectral sensitivities and the number of channels. Experiments are performed on simulations over hyperspectral data. The outcomes indicate that extension of computational color constancy algorithms from color to spectral gives promising results and may have the potential to lead towards efficient and stable representation across illuminants. However, this is highly dependent on spectral sensitivities and noise. We believe that the development of illuminant invariant multispectral imaging systems will be a key enabler for further use of this technology.

Color characterization methods for a multispectral camera

“…To overcome this limitation, the concept of multispectral constancy was recently proposed [38]. In this method, the estimation of illumination in multispectral images [39] is performed after the acquisition, and the effect of illumination is removed from the multispectral data. The spectral reconstruction system is trained/calibrated with the training data acquired with the same camera under a canonical illuminant.…”

Spectral Adaptation Transform for Multispectral Constancy

“…However, from a practical point of view, one reason for not employing such narrow band systems is that the acquisition time, complexity and cost of such system is high as compared to wide band systems [38]. Other reasons are the fact that the narrow band systems are not an optimal choice for illuminant estimation [16,17] and demosaicing (when multispectral filter array is used [39]). Wang et al [40] studied the influence of increase in number of bands and found that increasing the number of spectral bands cause reduction in performance of spectral reconstruction.…”

“…Wang et al [40] studied the influence of increase in number of bands and found that increasing the number of spectral bands cause reduction in performance of spectral reconstruction. Also, the efficiency of illuminant estimation algorithms decreases when the number of spectral filters is increased [17]. In this work, we are limiting the experiments to linear systems and our aim is to investigate the concept of multispectral constancy 13 for a generalized multispectral imaging system.…”

“…This transformation requires the estimation of scene illuminant. For multispectral images, estimation of illuminant in the sensor domain by using image statistics based methods is proposed by Thomas [16] and Khan et al [17]. In their work, the illuminant estimation methods for color images are extended from three channels to K channels.…”

Multispectral Constancy Based on Spectral Adaptation Transform

“…These algorithms are extensions of Max-RGB Algorithm [10] and the Gray-Edge Algorithm [11] [12]. The extension of these algorithms from color to spectral is discussed in detail in [9] [13]. Once the illuminant is estimated, then we propose to apply the SAT in form of a diagonal correction to the acquired data, so that it appears as if being taken under a canonical illuminant.…”