Color Models and Illumination Changes

Color model representation allows characterizing in a quantitative manner, any defined color spectrum of visible light, i.e. with a wavelength between 400nm and 700nm. To accomplish that, each model, or color space, is associated with a function that allows mapping the spectral power distribution of the visible electromagnetic radiation, in a space defined by a set of discrete values that quantify the color components composing the model. Some color spaces are sensitive to changes in lighting conditions. Others assure the preservation of certain chromatic features, remaining immune to these changes. Therefore, it becomes necessary to identify the strengths and weaknesses of each model in order to justify the adoption of color spaces in image processing and analysis techniques. This chapter will address the topic of digital imaging, main standards and formats. Next we will set the mathematical model of the image acquisition sensor response, which enables assessment of the various color spaces, with the aim of determining their invariance to illumination changes.

Eigenvalue computations in the context of data-sparse approximations of integral operators

In this work, we consider the numerical solution of a large eigenvalue problem resulting from a finite rank discretization of an integral operator. We are interested in computing a few eigenpairs, with an iterative method, so a matrix representation that allows for fast matrix-vector products is required. Hierarchical matrices are appropriate for this setting, and also provide cheap LU decompositions required in the spectral transformation technique. We illustrate the use of freely available software tools to address the problem, in particular SLEPc for the eigensolvers and HLib for the construction of H-matrices. The numerical tests are performed using an astrophysics application. Results show the benefits of the data-sparse representation compared to standard storage schemes, in terms of computational cost as well as memory requirements.

Error bounds for low-rank approximations of the first exponential integral kernel

A hierarchical matrix is an efficient data-sparse representation of a matrix, especially useful for large dimensional problems. It consists of low-rank subblocks leading to low memory requirements as well as inexpensive computational costs. In this work, we discuss the use of the hierarchical matrix technique in the numerical solution of a large scale eigenvalue problem arising from a finite rank discretization of an integral operator. The operator is of convolution type, it is defined through the first exponential-integral function and, hence, it is weakly singular. We develop analytical expressions for the approximate degenerate kernels and deduce error upper bounds for these approximations. Some computational results illustrating the efficiency and robustness of the approach are presented.