Color transform to optimize fruit ripeness discrimination in dichromats
Author
Antela, K.; Morales-Rubio, A.; Besada, Cristina; Tarancón, Paula; Cervera, M. L.; Luque, M. J.Date
2020Cita bibliográfica
Antela, K., Morales-Rubio, A., Besada, C., Tarancón, P., Cervera, M.L., & Luque, M.J. (2020). Color transform to optimize fruit ripeness discrimination in dichromats. En: Gene, A., Luque, MJ., Sañudo, F., Bueno, I., Herández, R., García, MC., Esteve, J. & Díez, MA. (Eds). V Congreso Internacional de Jóvenes Optometristas. pp: 83-84.Abstract
Aim: To develop and test a color transform for red-green color
defectives to enhance tomatoipeness judgements.
Experimental Method: Congenital protan and deutan color
defectives suffer sensitivity losses in the red-green mechanism [1]
compromising performance in color-discrimination-based everyday
tasks [2], which may be compensated by procedures designed to
optimize image color gamuts to minimize color confusion [3]. Given that
red-green defectives retain normal discrimination along the blue-yellow
axis in color space [1], we propose a simple procedure to recode redgreen
color differences in CIELAB color space as blue-yellow color
differences, to allow red-green defectives to correctly judge the ripeness
of tomatoes.
An agricultural cooperative of Perelló supplied and classified by
color the tomato samples in a controlled manner in four standard
ripeness stages. Sample color was measured with a portable Minolta
CR-300 colorimeter (Minolta Co. Ltd, Osaka, Japan). Tomatoes were
photographed with a Smartphone (Samsung Galaxy S7 edge model SMG935F
with a 12.2 MP camera). RGB values of the image were
transformed to XYZ values using Matlab’s sRGB transform, and then to
CIEL*a*b*, using as reference white a white sample illuminated as the
samples. Dichromatic perception of the images was simulated by the
corresponding pair algorithm [4]. The modified palette was obtained by
exchanging the values of the red-green (a*) and blue-yellow (b*)
descriptors (Fig. 1).