Generalization of the white-anchored correction factor

Determine whether the white-anchored correction factor of 0.74, estimated under D65 adaptation on the evaluated scene sets, remains valid across independent hyperspectral scene distributions and preserves the stated adequacy band.

Background

The adapted three-channel projection exhibits a residual common slope of approximately 0.74 rather than unit slope. The paper proposes the one-parameter white-anchored transformation y' = 1 + (y - 1)/0.74 as a possible correction that preserves the normalized white point.

However, the factor 0.74 was obtained from the current D65-adapted scene sets, and its applicability to other scene distributions has not been established. The paper specifies that the correction would fail if independent hyperspectral imagery showed a residual outside the adequacy band or if the common slope differed substantially from 0.74.

References

The correction factor $0.74$ was obtained under D65 adaptation on this scene set, however; whether the same factor holds across independent scene distributions (generalization) is untested.

— When is a closed-form RGB->S/P ratio adequate? A hyperspectral characterization on natural scenes for mesopic display  (2609.24819 - Uchida, 21 Sep 2026) in Section 6, item 'Recovery by a white-anchored transform (future test)'

The open work is an observer-validated evaluation of the transform it feeds and application to broader independent HSI (ICVL, KAUST-HS), against which the pre-registered band and the named failure class keep the result falsifiable.

— When is a closed-form RGB->S/P ratio adequate? A hyperspectral characterization on natural scenes for mesopic display  (2609.24819 - Uchida, 21 Sep 2026) in Section 6, final paragraph and item C1 'adequacy + generalization'