ix congreso nacional del color - Universidade do Minho

2 downloads 0 Views 5MB Size Report
Luís Jiménez del Barco Jaldo. Julio Antonio Lillo Jover. Francisco M. Martínez Verdú. Manuel Melgosa Latorre. Ángel Ignacio Negueruela. Susana Otero Belmar.
IX CONGRESO NACIONAL DEL COLOR ALICANTE 2010

SEDOPTICA

S O C I E D A D E S PA Ñ O L A D E Ó P T I C A

COMITÉ

E S PA Ñ O L

DE

COLOR

PUBLICACIONES UNIVERSIDAD DE ALICANTE

www.sri.ua.es/congresos/color10

Alicante, 29 y 30 de Junio, 1 y 2 de Julio de 2010 Universidad de Alicante

Este libro ha sido debidamente examinado y valorado por evaluadores ajenos a la Universidad de Alicante, con el fin de garantizar la calidad científica del mismo.

Publicaciones de la Universidad de Alicante Campus de San Vicente s/n 03690 San Vicente del Raspeig [email protected] http://publicaciones.ua.es Teléfono: 965903480 Fax: 965909445

© Varios autores, 2010 © de la presente edición: Universidad de Alicante

ISBN: 978-84-9717-144-1

Diseño de portada: candelaInk

Reservados todos los derechos. Cualquier forma de reproducción, distribución, comunicación pública o transformación de esta obra sólo puede ser realizada con la autorización de sus titulares, salvo excepción prevista por la ley. Diríjase a CEDRO (Centro Español de Derechos Reprográficos, www.cedro.org) si necesita fotocopiar o escanear algún fragmento de esta obra.

IX CONGRESO NACIONAL DEL COLOR. ALICANTE 2010

El IX Congreso Nacional de Color cuenta con el apoyo de las siguientes entidades:

IX CNC -Libro de Actas-

IX CONGRESO NACIONAL DEL COLOR. ALICANTE 2010

IX Congreso Nacional de Color Alicante, 29 y 30 de Junio, 1 y 2 de Julio Universidad de Alicante

Departamento de Óptica, Farmacología y Anatomía Facultad de Ciencias Instituto Universitario de Física Aplicada a las Ciencias y las Tecnologías (IUFACyT) Universidad de Alicante IX CNC -Libro de Actas-

IX CONGRESO NACIONAL DEL COLOR. ALICANTE 2010

COMITÉ ORGANIZADOR Presidente Vicepresidente I

Francisco M. Martínez Verdú Universidad de Alicante Universidad Politécnica de Valencia Eduardo Gilabert Pérez

Vicepresidente II

Joaquín Campos Acosta

IFA-CSIC

Secretaria Científica

Esther Perales Romero

Universidad de Alicante

Secretaria Administrativa

Olimpia Mas Martínez

Universidad de Alicante

Secretaria Técnica

Sabrina Dal Pont

Universidad de Alicante

Tesorero

Valentín Viqueira Pérez

Universidad de Alicante

Vocal

Elísabet Chorro Calderón

Universidad de Alicante

Vocal

Verónica Marchante

Universidad de Alicante

Vocal

Bárbara Micó Vicent

Universidad de Alicante

Vocal

Elena Marchante

Universidad de Alicante

Vocal

Ernesto R. Baena Murillo

Universidad de Alicante

COMITÉ CIENTÍFICO Natividad Alcón Gargallo

Instituto de Óptica, Color e Imagen, AIDO

Joaquín Campos Acosta

Instituto de Física Aplicada CSIC

Pascual Capilla Perea

Universidad de Valencia

Ángela García Codoner

Universidad Politécnica de Valencia

Eduardo Gilabert Pérez

Universidad Politécnica de Valencia

José Mª González Cuasante

Universidad Complutense de Madrid

Francisco José Heredia Mira

Universidad de Sevilla

Enrique Hita Villaverde

Universidad de Granada

Luís Jiménez del Barco Jaldo

Universidad de Granada

Julio Antonio Lillo Jover

Universidad Complutense de Madrid

Francisco M. Martínez Verdú

Universidad de Alicante

Manuel Melgosa Latorre

Universidad de Granada

Ángel Ignacio Negueruela

Universidad de Zaragoza

Susana Otero Belmar

Instituto de Óptica, Color e Imagen, AIDO

Jaume Pujol Ramo

Universidad Politécnica de Cataluña

Javier Romero Mora

Universidad de Granada

Mª Isabel Suero López

Universidad de Extremadura

Meritxell Vilaseca Ricart

Universidad Politécnica de Cataluña

IX CNC -Libro de Actas-

IX CONGRESO NACIONAL DEL COLOR. ALICANTE 2010

THE NUMBER OF DISCERNIBLE COLOURS PERCEIVED BY DICHOMATS IN NATURAL SCENES: THE INFLUENCE OF DIFFERENT THEORETICAL MODELS 1

João MM Linhares1, Sérgio MC Nascimento1 University of Minho, Department of Physics, Gualtar Campus, 4710-057, Braga, Portugal. [email protected]

Abstract: Several models are available to simulate the colour perception of dichromat observers. The goal of this work was to study how much the estimated number of discernible colours in natural scenes depends on the theoretical colour model applied. Twelve images of natural scenes were simulated for trichromats as if they were seen by dichromats. Two theoretical models were used in these simulations. The CIELAB colour volume was estimated to each image for normal observers and dichromats and for each of the models tested. The number of discernible colours was estimated by segmenting the CIELAB colour volume into unitary cubes and by counting the number of non-empty cubes, that is, assuming that all the colours that are inside the same cube could not be discernible. Reductions to less than 10% of the colours perceived by normal trichromats were found to all dichromats regardless the dichromatic colour model used. This result seems to indicate that the estimated chromatic diversity of natural scenes perceived by dichromats is not strongly influenced by the dichromatic colour model used. Key words: Colour vision deficiencies, Dichromats, Corresponding-pair procedure, Discernible colours, Natural scenes INTRODUCTION The chromatic diversity of complex scenarios can be used to estimate optimum illumination conditions for different visual tasks [1-3] or the influence of coloured filters[4]. The use of models that simulate for normal observers the colour perception of dichromats [5-6] enables the study of such diversity not only for normal colour vision observers but also for colour deficient observers. The goal of this work was to study how much the estimated number of discernible colours in natural scenes depends on the theoretical colour model applied. Twelve hyperspectral images of natural scenes were analysed and two dichromatic colour vision simulation models were used. Reductions to less than 10% of the colours perceived by normal trichromats were found for all dichromats regardless the dichromatic colour model used.

METHODS Figure 1 represents the thumbnails of an image database with hyperspectral data from 12 images of natural scenarios used in this work. Scenes of rural and urban environments were used [7-8]. The hyperspectral images were obtained with a hyperspectral imaging system with a low noise cooled digital camera (Hamamatsu, model C4742-95-12ER, Hamamatsu Photonics K. K., Japan), and a fast tunable liquid-crystal filter (Varispec, model VS-VIS2-10-HC-35-SQ, Cambridge Research & Instrumentation, Inc., Massachusetts) mounted in front of a lens, with an infrared blocking filter. The camera was capable of a spatial resolution of 1344 x 1024 pixels and each image was acquired from 400-720 nm in 10 nm steps. The spectrum of the light reflected from a gray surface present in the scene measure with a telespectroradiometer (SpectraColorimeter, PR-650, PhotoResearch Inc., Chatsworth, CA) just after image acquisition 87

IX CONGRESO NACIONAL DEL COLOR. ALICANTE 2010

was used to calibrate the hyperspectral data. The spectral radiance from each pixel of the image was then obtained after corrections for dark noise, spatial non-uniformities, stray light, and chromatic aberrations [7, 9]. Figure 2 represents the locus of the CIE1931 (x,y) chromaticity coordinates of the daylight illuminants of the analysed scenes.

Figure 1. Thumbnails of the 12 natural scenes analysed in this study.

For normal observers the spectral radiances were converted into tristimulus values for the CIE 1931 Standard Colorimetric Observer and then converted into the CIELAB colour coordinates [10]. Figure 3 represents the CIELAB colour volume computed for a normal observer for the natural scenes depicted. The reference illuminant for these computations was obtained from the gray reference surface present in the scene and the white object was assumed the perfect reflecting diffuser.

Figure 2. Locus of the CIE1931 (x,y) chromaticity coordinates of the daylight illuminant for each used scene.

Figure 3. CIELAB colour volume (gray symbols) and correspondent projection on the CIE (a*,b*) plane (light gray symbols) for the natural scene depicted.

For dichromats two models of colour vision were used to simulate for normal observers the colour stimulus experienced by dichromats [1, 5-6]. The Brettel et al. algorithm [5] converted the normal observer tristimulus values obtained from the radiance data into the (L,M,S) cone signals experienced by dichromats. In Capilla et al. algorith the Boynton model was used as a oneopponent-stage with no adaptation linear model and converted the normal observer tristimulus into dichromatic cone signals assuming the nulling hypothesis and the S cone equals to the M cone; the ATD95 model was used as a two-opponent-stage linear model [6] assuming the observer adapted to a dark background and a second-stage nulling for protanopes and deuteranopes and a first-stage nulling for tritanopes and converted the normal observer tristimulus into dichromatic cone signals assuming the conversion of the tristimulus values units into trolands, the Smith and Pokorny fundamentals [11] with the L,M,S sensitivity curves 88

IX CONGRESO NACIONAL DEL COLOR. ALICANTE 2010

normalized to 0.66:1:0.43, respectively, and a gain control in the non-linearity stage of 300. The use of an inversion model adapted for each case enabled the conversion of the dichromatic cone signals into tristimulus values experienced by normal observers which enables the conversion of such stimulus into the CIELAB colour coordinates. All the colours converted and analysed had a real correspondent pair. The Boynton and the ATD95 models were used in the Capilla et al. algorithm as they correspond to the farther apart results when compared to the Brettel et al. algorithm (as represented in Fig 4 of Capilla et al. paper [6]).

ATD95

Boynton

Brettel et al.

Model

Normal Observer

Protanope observer

Deuteranope observer

Tritanope observer

Figure 4. Natural scene image as .seen by a normal, a protanope, a deuteranope and a tritanope observer as predicted by the Brettel et al, Boynton (nulling hypothesis, with S=M for the tritanope) and ATD95 (second stage nulling for protanopes and deuteranopes and first stage nulling with S=M for tritanopes) models.

The number of discernible colours was estimated by segmenting the CIELAB colour volume into unitary cubes and by counting the number of non-empty cubes. It was assumed that all the colours that are inside the same cube could not be discernible. The percentage of the colours perceived by dichromats when compared to a normal observer was used to compare the different models rather than absolute values of the number of discernible colours.

RESULTS Figure 3 represents the images simulated for normal observers as if they were seen by dichromats using the Brettel et al. and Capilla et al for Boynton and ATD95 models colour vision models. Figure 4 represents the CIELAB colour volumes (gray symbols) and correspondent projection on the CIE (a*,b*) plane (light gray symbols) obtained using the different colour simulation models for dichromats. Considerable reductions in the CIELAB colour volume were observed for each dichromatic observer when compared to a normal observer (as represented in Figure 5) despite the colour simulation model used. 89

IX CONGRESO NACIONAL DEL COLOR. ALICANTE 2010

Deuteranope

Tritanope

ATD95

Boynton

Brettel et al.

Protanope

Model

Dichromatic observer

Figure 5. CIELAB colour volumes (gray symbols) and correspondent projection on the CIE (a*,b*) plane (light gray symbols) for the scene represented in Figure 2 for a normal, a protanope, a deuteranope and a tritanope observer as predicted by the Brettel et al, Boynton (nulling hypothesis, with S=M for the tritanope) and ATD95 (second stage nulling for protanopes and deuteranopes and first stage nulling with S=M for tritanopes) models. The CIELAB colour volume is scale as the normal colour volume in Figure 2. Table 1. Percentage of the colours perceived by dichromats when compared to a normal observer as predicted by the Brettel et al, Boynton (nulling hypothesis, with S=M for the tritanope) and ATD95 (second stage nulling for protanopes and deuteranopes and first stage nulling with S=M for tritanopes) models. Data represents the average across scenes for each type of observer and in brackets the standard deviation (STD). Model

Protanope

Deuteranope

Tritanope

Brettel et al.

6(±2)%

7(±2)%

7(±2)%

Boynton

8(±2)%

8(±2)%

6(±2)%

ATD95

6(±2)%

7(±2)%

9(±3)%

Table 1 shows the percentage of discernible colours perceived by dichromats when compared to normal observers as a function of the used simulation model, and in all cases less than 10%. Data was obtained by estimating the arithmetic average of the impairment of the number of discernible colours across scenes with associated Standard Deviation error. 90

IX CONGRESO NACIONAL DEL COLOR. ALICANTE 2010

CONCLUSIONS In this work we estimated the chromatic diversity impairment experienced by dichromats when viewing hyperspectral images of natural scenes using different colour simulation models. It was found that the impairment to all dichromats is greater than 90% when compared to normal observers. Such result is similar to the ones found in other studies [1, 4] and seems to indicate that the model used to simulate the chromatic diversity that dichromats perceived doesn’t strongly influence the final result. The models applied have known limitations [6] and the influence of other adaptation backgrounds than the black was not estimated. The colour space used to estimate the number of discernible colours is also known for its non-uniformity [12]. Despite these limitations the data presented here seems to indicate that the chromatic diversity of natural scenes perceived by dichromats is not strongly influenced by the dichromatic colour simulation model used.

ACKNOWLEDGEMENTS This work was supported by the Centro de Física of Minho University, Braga, Portugal, by the Fundação para a Ciência e a Tecnologia (grants POSC/EEA-SRI/57554/2004 and POCTI/EAT/55416/2004). João M.M. Linhares was supported by grant SFRH / BD / 35874 / 2007. A special thanks to María José Luque and Esther Perales Romero for their comments and help on the corresponding-pair procedure algorithm.

REFERENCES [1]

Linhares, J.M.M., P.D.A. Pinto, and S.M.C. Nascimento, "Color rendering of art paintings under CIE illuminants for normal and color deficient observers". Journal of the Optical Society of America a-Optics Image Science and Vision, Vol. 26, No. 7, pág. 1668-1677.(2009) [2] Pinto, P.D., et al., "Psychophysical estimation of the best illumination for appreciation of Renaissance paintings". Vis Neurosci, Vol. 23, No. 3-4, pág. 669-674.(2006) [3] Pinto, P.D., J.M.M. Linhares, and S.M.C. Nascimento, "Correlated color temperature preferred by observers for illumination of artistic paintings". Journal of the Optical Society of America A: Optics, Image Science, and Vision, Vol. 25, No. 3, pág. 623-630.(2008) [4] Linhares, J.M.M., P.D. Pinto, and S.M.C. Nascimento, "The number of discernible colors perceived by dichromats in natural scenes and the effects of colored lenses". Visual Neuroscience, Vol. 25, No. 3, pág. 493499.(2008) [5] Brettel, H., F. Viénot, and J.D. Mollon, "Computerized simulation of color appearance for dichromats". JOSAA, Vol. 14, No. 10, pág. 2647-2655.(1997) [6] Capilla, P., et al., "Corresponding-pair procedure: a new approach to simulation of dichromatic color perception". Journal of the Optical Society of America a-Optics Image Science and Vision, Vol. 21, No. 2, pág. 176-186.(2004) [7] Foster, D.H., et al., "Frequency of metamerism in natural scenes". Journal of the Optical Society of America A: Optics, Image Science, and Vision, Vol. 23, No. 10, pág. 2359-2372.(2006) [8] Foster, D.H., S.M.C. Nascimento, and K. Amano, "Information limits on neural identification of colored surfaces in natural scenes". Visual Neuroscience, Vol. 21, No. 3, pág. 331-336.(2004) [9] Linhares, J.M., P.D. Pinto, and S.M. Nascimento, "The number of discernible colors in natural scenes". J Opt Soc Am A Opt Image Sci Vis, Vol. 25, No. 12, pág. 2918-2924.(2008) [10] CIE, Colorimetry, CIE Publ 15:2004(CIE: Viena, 2004). [11] Smith, V.C. and J. Pokorny, "Spectral Sensitivity of Foveal Cone Photopigments between 400 and 500 Nm". Vision Research, Vol. 15, No. 2, pág. 161-171.(1975) [12] Luo, M.R., G. Cui, and B. Rigg, "The development of the CIE 2000 colour-difference formula: CIEDE2000". Color Research and Application, Vol. 26, No. 5, pág. 340-350.(2001)

91