Psychophysical Determination of the Relevant Colours That Describe the Colour Palette of Paintings
"> Figure 1
<p>Paintings used in the experiment from (uppermost eight rows) the Prado museum [<a href="#B20-jimaging-07-00072" class="html-bibr">20</a>] and from (lowermost eight rows) the Khan et al. dataset [<a href="#B22-jimaging-07-00072" class="html-bibr">22</a>].</p> "> Figure 2
<p>(<b>a</b>) Example of the colours selected as relevant colours (magenta circles) for one image from the Prado museum (<span class="html-italic">The Crucifixion</span>, by Juan de Flandes, 1509–1519) and one example from the Khan database (<span class="html-italic">Seated woman with wrist watch</span>, by Picasso, 1932). (<b>b</b>) CIEa*b* colour coordinates of all the colours within the image and the relevant colours chosen by one observer according to the subjective impression of the corresponding painting.</p> "> Figure 3
<p>Distribution of hue versus chroma values for all the original colours and the subjective relevant colours selected by one observer for (<b>a</b>) <span class="html-italic">The Crucifixion</span>, and (<b>b</b>) <span class="html-italic">Seated woman with wrist watch</span>.</p> "> Figure 4
<p>Average numbers of subjective relevant colours obtained for all painting images and all observers.</p> "> Figure 5
<p>All computational (blue points) and subjective (magenta circles) relevant colours obtained from all observers. The upper row corresponds to the Prado museum paintings, and the lower row corresponds to the Khan image dataset.</p> "> Figure 6
<p>Comparison between the computational colours and relevant colours obtained for the painting <span class="html-italic">Queen Joanna the Mad</span>, by Pradilla, 1877 (which corresponds to image number 19, fourth row, in <a href="#jimaging-07-00072-f001" class="html-fig">Figure 1</a>) and one of the observers. Computational colours are marked as blue points, and the obtained experimental relevant colours are marked as magenta circles.</p> "> Figure 7
<p>(<b>a</b>) Original painting and (<b>c</b>) its corresponding segmented counterpart based on the subjective relevant colours describing the colour palette of the painting (<b>b</b>). The numbers of relevant colours in each palette were 29, 41, 16, and 32, respectively, from the top to bottom images. The correlation coefficients between each original image and each segmented image were 0.9554, 0.9718, 0.9148, and 0.9249, respectively.</p> "> Figure 8
<p>(<b>a</b>,<b>b</b>) Hue versus chroma colour components derived from the subjective relevant colours for all observers. (<b>c</b>) Polar histogram of the frequency distribution of the hue and (<b>d</b>) chroma colour components obtained for both sets of paintings and all observers.</p> "> Figure 9
<p>(<b>a</b>) Pseudo-colour representation of the spatio-chromatic structure of the activation for all 192 learned independent component analysis (ICA) basis functions. Each colour component corresponds to the R, G, and B pixel values for each 7 × 7 patch that was derived from individual basis functions. (<b>b</b>) Chromaticities of activation of each individual patch when projected onto a red-green and blue-yellow plane in the RGB colour space. The functions are in order of decreasing <span class="html-italic">L</span><sup>2</sup> norm from left to right and top to bottom.</p> "> Figure 10
<p>(<b>a</b>) Histogram of the distribution of chromatic directions associated to each ICA basis function using the Khan set of painting images. (<b>b</b>) Histogram of colour angles in the RG and BY colour plane obtained from the relevant colours and all observers used in the psychophysical experiment.</p> ">
Abstract
:1. Introduction
2. Materials and Methods
2.1. Image Datasets
2.2. Procedure
2.3. Observers
2.4. Statistical Descriptors of the Paintings
- Self-similarity: by using the histogram intersection kernel and comparing the histogram of oriented gradients (HOG) features of each sub-image at level 3 with those of the entire image at level 0;
- Complexity: computed as the mean norm of the gradient across all orientations over the gradient image;
- Birkhoff-like metric: computed as the ratio between the self-similarity and complexity metrics;
- Anisotropy: calculated as the variance of all the HOG values at level 3.
3. Results
3.1. Subjective Relevant Colours versus Computational Relevant Colours
3.2. Colour Palette of Paintings Estimated from the Subjective Relevant Colours
3.3. Subjective Relevant Colours, Colour Features, and Efficient Coding
4. Discussion and Conclusions
Supplementary Materials
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Acknowledgments
Conflicts of Interest
References
- Graham, D.J.; Redies, C. Statistical regularities in art: Relations with visual coding and perception. Vis. Res. 2010, 50, 1503–1509. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Wallraven, C.; Fleming, R.; Cunningham, D.; Rigau, J.; Feixas, M.; Sbert, M. Categorizing art: Comparing humans and computers. Comput. Graph. 2009, 33, 484–495. [Google Scholar] [CrossRef]
- Mureika, J.R.; Dyer, C.C.; Cupchik, G.C. Multifractal structure in nonrepresentational art. Phys. Rev. E 2005, 72, 046101. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Yelizaveta, M.; Tat-Seng, C.; Irina, A. Analysis and retrieval of paintings using artistic color concepts. In Proceedings of the 2005 IEEE International Conference on Multimedia and Expo, Amsterdam, The Netherlands, 6 July 2005; pp. 1246–1249. [Google Scholar]
- Montagner, C.; Linhares, J.M.; Vilarigues, M.; Nascimento, S.M. Statistics of colors in paintings and natural scenes. JOSA A 2016, 33, A170–A177. [Google Scholar] [CrossRef]
- Nascimento, S.M.; Linhares, J.M.; Montagner, C.; João, C.A.; Amano, K.; Alfaro, C.; Bailão, A. The colors of paintings and viewers’ preferences. Vis. Res 2017, 130, 76–84. [Google Scholar] [CrossRef] [PubMed]
- Romero, J.; Gómez-Robledo, L.; Nieves, J. Computational color analysis of paintings for different artists of the XVI and XVII centuries. Color Res. Appl. 2018, 43, 296–303. [Google Scholar] [CrossRef]
- Li, C.; Chen, T. Aesthetic visual quality assessment of paintings. IEEE J. Sel. Top. Signal Process. 2009, 3, 236–252. [Google Scholar] [CrossRef]
- Frey, H.; Honey, C.; König, P. What’s color got to do with it? The influence of color on visual attention in different categories. J. Vis. 2008, 8, 6. [Google Scholar] [CrossRef]
- Heckbert, P. Color image quantization for frame buffer display. Siggraph Comput. Graph. 1982, 16, 297–307. [Google Scholar] [CrossRef] [Green Version]
- Google Arts & Culture. Available online: https://artsandculture.google.com/ (accessed on 2 February 2018).
- Colormind. Available online: http://colormind.io/ (accessed on 2 February 2018).
- Simoncelli, E.P.; Olshausen, B.A. Natural image statistics and neural representation. Annu. Rev. Neurosci. 2001, 24, 1193–1216. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Graham, D.J.; Field, D.J. Efficient neural coding of natural images. New Encycl. Neurosci. 2007, 1, 1–18. [Google Scholar]
- Linhares, J.M.M.; Pinto, P.D.; Nascimento, S.M.C. The number of discernible colors in natural scenes. JOSA A 2008, 25, 2918–2924. [Google Scholar] [CrossRef]
- Masaoka, K.; Berns, R.S.; Fairchild, M.D.; Abed, F.M. Number of discernible object colors is a conundrum. JOSA A 2013, 30, 264–277. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Witzel, C.; Gegenfurtner, K.R. Color perception: Objects, constancy, and categories. Annu. Rev. Vis. Sci. 2018, 4, 475–499. [Google Scholar] [CrossRef]
- Párraga, C.; Benavente, R.; Baldrich, R.; Vanrell, M. Psychophysical measurements to model intercolor regions of color-naming space. J. Imaging Sci. Technol. 2009, 53, 31106–31108. [Google Scholar] [CrossRef] [Green Version]
- Yu, L.; Zhang, L.; van de Weijer, J.; Khan, F.S.; Cheng, Y.; Parraga, C.A. Beyond eleven color names for image understanding. Mach. Vis. Appl. 2018, 29, 361–373. [Google Scholar] [CrossRef]
- Nieves, J.L.; Gomez-Robledo, L.; Chen, Y.; Romero, J. Computing the relevant colors that describe the color palette of paintings. Appl. Opt. 2020, 59, 1732–1740. [Google Scholar] [CrossRef] [PubMed]
- Prado Museum. Available online: https://www.museodelprado.es/en/the-collection (accessed on 14 April 2021). (only Spanish version).
- Khan, F.S.; Beigpour, S.; Van de Weijer, J.; Felsberg, M. Painting-91: A large scale database for computational painting categorization. Mach. Vis. Appl. 2014, 25, 1385–1397. [Google Scholar] [CrossRef] [Green Version]
- Psychtoolbox 3. Available online: http://psychtoolbox.org/ (accessed on 15 June 2019).
- Eizo Color Edge CG277. Available online: https://eizo.es/producto/cg277-coloredge/ (accessed on 15 May 2019).
- Nodine, C.; Mello-Thoms, C.; Krupinski, E.; Locher, P. Visual interest in pictorial art during an aesthetic experience. Spat Vis. 2008, 21, 55–77. [Google Scholar] [CrossRef]
- Van Geert, E.; Wagemans, J. Order, complexity, and aesthetic appreciation. Psychol. Aesthet. Creat. Arts 2020, 14, 135. [Google Scholar] [CrossRef] [Green Version]
- Hugo GävertJarmo HurriJaakko SäreläAapo Hyvärinen, FastICA Matlab Package. Available online: http://research.ics.aalto.fi/ica/fastica/ (accessed on 1 September 2020).
- Nieves, J.L.; Romero, J. Heuristic analysis influence of saliency in the color diversity of natural images. Color Res. Appl. 2018, 43, 713–725. [Google Scholar] [CrossRef]
- Barlow, H.B. Possible principles underlying the transformation of sensory messages. Sens. Commun. 1961, 1, 217–234. [Google Scholar]
- Bell, A.J.; Sejnowski, T.J. The “independent components” of natural scenes are edge filters. Vis. Res. 1997, 37, 3327–3338. [Google Scholar] [CrossRef] [Green Version]
- Nadal, J.; Parga, N. Information processing by a perceptron in an unsupervised learning task. Netw. Comput. Neural Syst. 1993, 4, 295–312. [Google Scholar] [CrossRef]
- Wachtler, T.; Lee, T.; Sejnowski, T.J. Chromatic structure of natural scenes. JOSA A 2001, 18, 65–77. [Google Scholar] [CrossRef] [PubMed]
- Kellner, C.J.; Wachtler, T. A distributed code for color in natural scenes derived from center-surround filtered cone signals. Front. Psychol. 2013, 4, 661. [Google Scholar] [CrossRef] [Green Version]
- Ruderman, D.L.; Cronin, T.W.; Chiao, C. Statistics of cone responses to natural images: Implications for visual coding. JOSA A 1998, 15, 2036–2045. [Google Scholar] [CrossRef] [Green Version]
- Hasantash, M.; Afraz, A. Richer color vocabulary is associated with better color memory but not color perception. Proc. Natl. Acad. Sci. USA 2020, 117, 31046–31052. [Google Scholar] [CrossRef] [PubMed]
- Bartleson, C.J. Memory Colors of Familiar Objects*. J. Opt. Soc. Am. 1960, 50, 73–77. Available online: http://www.osapublishing.org/abstract.cfm?URI=josa-50-1-73 (accessed on 15 February 2021). [CrossRef] [PubMed]
- Hanley, J.R. Color categorical perception. Encycl. Color Sci. Technol. 2015, 1–6. [Google Scholar] [CrossRef]
- Foster, D.H. The Verriest Lecture: Color vision in an uncertain world. JOSA A 2018, 35, B192–B201. [Google Scholar] [CrossRef] [PubMed]
- Zhaoping, L. Theoretical understanding of the early visual processes by data compression and data selection. Netw. Comput. Neural Syst. 2006, 17, 301–334. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Fontoura, P.; Menu, M.; Schaeffer, J.M. Visual perception of natural colors in paintings: An eye-tracking study of Grünewald’s Resurrection. Color Res. Appl. 2021, 46, 582–594. [Google Scholar] [CrossRef]
- Patterson, S.S.; Neitz, M.; Neitz, J. Reconciling color vision models with midget ganglion cell receptive fields. Front. Neurosci. 2019, 13, 865. [Google Scholar] [CrossRef] [PubMed]
Set | NDC | NRC | L* | a* | b* | Angle (°) | Ratio | Area | Self-Similarity | Complexity | Anisotropy | Birkhoff Metric | Volume |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1 | 56,754 | 19 | 35.0 | 4.2 | 12.9 | 67 | 0.41 | 2011 | 0.685 | 2.5 | 8.47 × 10−7 | 0.283 | 142,216 |
SD | 5 | 9.8 | 3.7 | 4.3 | 17 | 0.11 | 1431 | 0.085 | 0.6 | 3.03 × 10−7 | 0.051 | 59,450 | |
2 | 77,125 | 21 | 56.3 | 2.9 | 14.3 | 74 | 0.49 | 7528 | 0.612 | 4.4 | 1.62 × 10−6 | 0.190 | 337,709 |
SD | 7 | 16.1 | 8.5 | 12.2 | 26 | 0.15 | 6923 | 0.246 | 3.2 | 1.21 × 10−6 | 0.156 | 141,009 |
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |
© 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Nieves, J.L.; Ojeda, J.; Gómez-Robledo, L.; Romero, J. Psychophysical Determination of the Relevant Colours That Describe the Colour Palette of Paintings. J. Imaging 2021, 7, 72. https://doi.org/10.3390/jimaging7040072
Nieves JL, Ojeda J, Gómez-Robledo L, Romero J. Psychophysical Determination of the Relevant Colours That Describe the Colour Palette of Paintings. Journal of Imaging. 2021; 7(4):72. https://doi.org/10.3390/jimaging7040072
Chicago/Turabian StyleNieves, Juan Luis, Juan Ojeda, Luis Gómez-Robledo, and Javier Romero. 2021. "Psychophysical Determination of the Relevant Colours That Describe the Colour Palette of Paintings" Journal of Imaging 7, no. 4: 72. https://doi.org/10.3390/jimaging7040072