ABSTRACT
Despite recent advances in digital color imaging [1], [2], producing digital photographs that are faithful representations of the original visual scene remains still challenging due to a number of constraints under which digital cameras operate. In particular, differences in characteristics between image acquisition systems and the human visual system constitute an underlying problem in digital color imaging. Digital cameras typically use various color filters [3], [4], [5] to reduce the available color information to the light of certain wavelengths, which is then sampled by the image sensor, such as a charge-coupled device
(CCD) [6], [7] or a complementary metal oxide semiconductor (CMOS) sensor [8], [9]. Unfortunately, achieving a precise digital representation of the visual scene is still not quite possible and extensive processing [10], [11], [12] of acquired data is needed to reduce an impact of shortcomings of the imaging system.
