AUTHOR=Zhang Yunhong , Hu Yuelin , Tan Jun , Ma Ruiqing , Si Feng , Yang Yi TITLE=Do color enhancement algorithms improve the experience of color-deficient people? An empirical study based on smartphones JOURNAL=Frontiers in Neuroscience VOLUME=18 YEAR=2024 URL=https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2024.1366541 DOI=10.3389/fnins.2024.1366541 ISSN=1662-453X ABSTRACT=

Approximately 8% of the global population experiences color-vision deficiency. It is important to note that “color-vision deficiency” is distinct from “color blindness,” as used in this article, which refers to the difficulty in distinguishing certain shades of color. This study explores color enhancement algorithms based on the neural mechanisms of color blindness and color deficiency. The algorithms are then applied to smartphones to improve the user experience (UX) of color-enhancing features in different top-selling smartphone brands with different operating systems (OS). A color-enhancing application program was developed for individuals with color-vision deficiency and compared to two other mature color-enhancing programs found in top-selling smartphones with different mainstream operating systems. The study included both objective and subjective evaluations. The research materials covered three aspects: daily life, information visualization, and videos. Additionally, this research study examines various levels of color enhancement through three dimensions of subjective evaluation: color contrast, color naturalness, and color preference. The results indicate that all color-enhancing features are beneficial for individuals with color-vision deficiencies due to their strong color contrast. The users' color preference is closely linked to color naturalness. The application program preserves the naturalness of colors better than the other two color-enhancing features. The subjective evaluations show similar trends across different operating systems, with differences arising from the use of different color-enhancing algorithms. Therefore, different algorithms may result in different sizes of the color gamut.