Color Management Using Optimal Three-Dimensional Look-Up Tables

Abstract

Digital imaging has seen an unprecedented growth in the past five years. The variety of imaging systems available to users to create and view visual data is enormous. Color management has become an important aspect of modern imaging and display systems. Color profiles have been the de facto tool for achieving faithful visual reproduction for a long time. In this article, the authors discuss issues associated with profile-based color management systems. The authors describe an alternative approach motivated by the problem of visually matching two known display devices. The authors use a model-based method to achieve this goal and propose realizing the method with simple table look-up operations. The authors devise a framework for designing look-up tables (LUTs) which are optimal in terms of color reproduction on the displays based on resource constraints. The LUT-based color management system is shown to be more accurate and memory-efficient than a comparable International Color Consortium profile-based system.

Image enhancement and quality measures for dietary assessment using mobile devices

“…From the results summarized in Table 1 and Figure 7, we can observe that our proposed methods usually have better color consistency in the various illumination tests than the methods derived by Choi[19] and Srivastava[20] in that the error, Δ, is smaller. Figure 7 shows the images after color correction.…”

“…The linear model and the nonlinear model in RGB color space are better than the methods described by Choi[19] and Srivastava[20]. They both perform similarly and improve the overall color consistency.…”

“…The captured images under non-D65 illumination were then color corrected using the three methods described in the previous section as well as two other methods described by Choi[19] and Srivastava[20]. To evaluate the accuracy of each method, the euclidean distance between the average color of each color patch on the GretagMacbeth Colorchecker $({\tilde{R}}_i^t,{\tilde{G}}_i^t,{\tilde{B}}_i^t)$ and the known sRGB values of each patch under D65 illumination $(R_i^r,G_i^r,B_i^r)$ (see Figure 6) was obtained using Equation 17 and are shown in Table 1.…”

“…We also adopted a simplified version of the approach proposed by Srivasrava, et al[20] to address the problem of visually matching two known display devices in color management systems. The method is implemented through the use of 3D look-up tables (LUT).…”

“…Sample color correction results using these methods, as well as comparison to the methods described by Choi[19] and Srivastava[20] are described next in Section 4.…”