A revised version of the psychophysical color difference Munsell Re-renotation dataset.
The revised datasets originate from papers:
[1] V. Timofeev et al., "No-Reference Error Detection in Color Difference Datasets: Application to Munsell Data," in IEEE Access, vol. 13, pp. 109322-109344, 2025, doi: 10.1109/ACCESS.2025.3582053.
[2] Nikolaev, D., Basova, O., Usaev, G., Tchobanou, M., & Bozhkova, V. (2023, June). Detection and Correction of Errors in Psychophysical Color Difference Munsell Re-renotation Dataset. In London Imaging Meeting (Vol. 4, pp. 40-44). Society for Imaging Science and Technology. https://doi.org/10.2352/lim.2023.4.1.10
Munsell Re-renotation is a psychophysical dataset describing large color differences, featuring 2986 colors characterized by standard colorimetric coordinates (x, y, Y) and coordinates within the Munsell system (H, V, C). The Munsell Re-renotation iteration enhances the uniformity of the system compared to its predecessor, the Munsell Renotation dataset.
The original Munsell Re-renotation data exists as a scanned report, accessible at http://www.rit-mcsl.org/MunsellRenotation/MunsellRe-renotations.pdf
Within this repository, you will find the Munsell Re-renotation data converted into a machine-readable format (.csv), accompanied by four revised versions:
- Corrected Misprints (35 items). This version rectifies 35 misprints identified during the initial round of corrections from the scanned report [2].
- Excluded Ambiguous and Duplicate Colors (5 Items). Ambiguous and duplicate colors, totaling 5 items, have been excluded from this version [2].
- Second Iteration Misprints Corrections (41 Items). Incorporating 41 additional misprint corrections identified during a subsequent review [1].
- Additional Misprints (10 Items). Incorporating additional misprint corrections identified during visual analysis of data in CIE xyY space. These corrections are not reflected in the papers [1, 2]; a list of changes can be found in the changelog/3.2-3.3.csv.
Additionally, the repository includes Munsell Renotation data sourced from https://www.rit.edu/science/munsell-color-science-lab-educational-resources (file real.dat), converted to .csv format for ease of use.
Repository Contents:
/munsell_2-0.csv: Munsell Renotation data containing real colors only, i.e., within the Macadam limits, sourced from the RIT website (file real.dat converted to .csv format)./munsell_3-0.csv: Original Munsell Re-renotation data from the report, featuring 2986 colors./munsell_3-1-0.csv: Munsell Re-renotation data with 35 corrected erroneous colors (2986 colors)./munsell_3-1-1.csv: Munsell Re-renotation data with 35 corrected erroneous colors, excluding 4 ambiguous colors and 1 duplicate color (2981 colors)./munsell_3-2.csv: Munsell Re-renotation data incorporating a total of 76 corrected erroneous colors from munsell_3-0.csv (2985 colors). Note that the corrections applied resulted in the removal of one color from the dataset./munsell_3-3.csv: Data incorporating 10 additional corrected erroneous colors, identified during visual analysis of data in CIE xyY space.changelog/3.3.csv: Description of additional errors found, not presented in papers [1, 2].
All data within this repository are available under the Creative Commons Attribution-NonCommercial-ShareAlike (CC BY-NC-SA 4.0) license. For more information, refer to the license at https://creativecommons.org/licenses/by-nc-sa/4.0.