A framework and methodology for spectral color vision deficiency imaging

Abstract

People with color vision deficiency (CVD) face difficulties in everyday life and may get frustrated when they miss some of the important features in an image because of their inability to perceive differences between certain colors that can be distinguished by normal color vision people. This can affect access to education and choice in their career. In order to help millions of people affected by color blindness worldwide, a method known as daltonization is used, which tries to modify colors in a photographic image in order to increase the color contrast and bring back the missing features, thus improving the accessibility of the images in terms of retrieving information content for CVD people. A daltonization algorithm to work successfully, accurate conversion of a color image to a corresponding CVD image is vital. Many CVD simulation methods have been proposed, and most of them rely on color appearance theories and models that are imperfect. In addition, these methods are based on a generalization of CVD models and types. Moreover, color imaging has well-known limitations like environment dependency, metamerism problem, and it is limited to the visual spectrum. Since spectral imaging addresses these limitations effectively, spectral imaging-based simulation of CVD can produce individualized and more accurate results. In this paper, we present a framework for spectral imagebased CVD imaging, which we call spectral CVD imaging, which can acquire an accurate personalized CVD image of a scene real-time under an uncontrolled illumination condition.