Accuracy Improvement of Estimated Perceived Brightness Maps by Helmholtz-Kohlrausch Effect Using a Correction Coefficient

Hashimoto, Shinichi; Shizume, Takaya; Takamatsu, Hiroaki; Shimodaira, Yoshifumi; Ohashi, Gosuke

doi:10.1587/transfun.e100.a.565

Cited by 3 publications

(1 citation statement)

References 2 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…7,8 We have proposed estimation equations of the H-K effect in natural images, which generally present various combinations of colors and pixels on display devices. [9][10][11][12][13] These natural images consist of multiple monochromatic images. However, our own estimation of the H-K effect treats a natural image as a set of monochromatic images.…”

Section: Introductionmentioning

confidence: 99%

Estimating the Helmholtz–Kohlrausch effect in natural images considering the decreasing brightness‐to‐luminance ratio with increasing lightness

et al. 2021

Self Cite

View full text Add to dashboard Cite

For colors with the same luminance, increasing the saturation increases the color brightness. This phenomenon, known as the Helmholtz-Kohlrausch (H-K) effect, is a characteristic of human vision. In our estimations of the H-K effect in natural images, we found that the brightness-to-luminance (B/L) ratio decreases with increasing the lightness. However, the estimation was not enough for high lightness images containing a large amount of cyan. We, therefore, propose an estimation equation considering the decreasing B/L ratio with increasing the lightness. This estimation method takes into account the characteristic that the B/L ratio is 1.0 or higher. First, we create a function that reflects the decrease of the magnitude of the H-K effect with respect to the increase of the lightness at each hue. Then, the difference between the lightness and the standard lightness (the lightness used in Nayatani's estimation equation) is multiplied. The effectiveness of the proposed method was verified in a subjective-evaluation experiment of high-lightness images containing a large amount of cyan. Moreover, the method improved the accuracy of the estimation, and we confirmed the decrease in B/L ratio with increasing the lightness by excluding B/L values below 1.0.

show abstract