Ana María Martín Casado

Comparison of the CIELab and CIEDE2000 color difference formulas.

STATEMENT OF PROBLEM Many color specification systems and color differences have been proposed to improve the correlation between color measurement and visual perception. Although color differences can be quantified using either the CIELab formula (ΔE*(ab)) or the recently introduced CIEDE2000 formula (ΔE00), which captures the perceived color difference better is unknown. PURPOSE The purpose of this study was to evaluate the CIELab and CIEDE2000 formulas to determine which best reflects the difference in color perception and whether color perception differs by sex. MATERIAL AND METHODS Forty participants grouped 18 dental resin disks (color range: from 73.6 to 87.5 for L*; from -1.6 to 3.4 for a*; from 18.1 to 36.6 for b*), the only requirement being that each group was formed of disks with chromatically indistinguishable colors. Each participant was free to choose the number and composition of the groups. With the results obtained, a dissimilarity matrix was generated, and nonmetric multidimensional scaling (MDS) was applied to it to obtain the coordinates of the disks within a Euclidean space. RESULTS The linear correlation coefficient between the interpoint distances of the MDS configuration (MDS_total sample) and the color differences with the CIELab formula (ΔE*(ab)) was 0.176 (P=.029), whereas with the CIEDE2000 formula (ΔE00) it was 0.289 (P<.001). For the configuration obtained in the womens group (MDS_w), the correlation between the interpoint distances and the CIELab color differences was 0.230 (P=.004), and the CIEDE2000 color difference was 0.328 (P<.001). For the configuration obtained in the mens group (MDS_m), the color differences calculated with both formulas reflected the perceived differences more poorly. CONCLUSIONS Within the limitations of this study, the CIEDE2000 formula reflected the color differences perceived by the human eye better than the CIELab formula (ΔE*ab). In addition, women were confirmed to be more sensitive than men to color differences.

Comparison of two color-difference formulas using the Bland-Altman approach based on natural tooth color space.

STATEMENT OF PROBLEM Several formulas to measure color differences have been proposed to improve the correlation between color measurements and visual perception. Despite the progress obtained, equating between formulas is complex, rendering the exchange of information in scientific papers difficult. PURPOSE The purpose of this study was to determine the relationship between the results provided by classic CIELab (ΔEab∗) and the CIEDE2000 (ΔE00) formulas in the natural tooth color space using the Bland and Altman limits of agreement, to use this relationship to establish the equivalences between the tooth color thresholds of acceptability and perceptibility of both formulas, and to evaluate whether the relationship between ΔEab∗ and ΔE00 is modified depending on the axis on which the changes occur. MATERIAL AND METHODS The L∗, a∗, and b∗ means coordinates in the 26 shade tabs of the Toothguide 3D Master were used, and color differences were calculated in 325 pairs of shade tabs using the CIELab (ΔE∗ab) and CIEDE2000 (ΔE00) color difference formulas. The results obtained with these formulas were compared, and the limits of agreement after a logarithmic transformation of the data were obtained. RESULTS The linear relationship between both formulas was ΔE00 =0.66ΔE∗ab. The results obtained with the CIELab formula were between 1.15 (95% confidence interval 1.11 to 1.18) and 2.09 (95% confidence interval 2.03 to 2.15) times higher than those obtained with the CIEDE2000 formula. CONCLUSIONS In the natural tooth color space, the scale factor between CIEDE2000 and CIELab values changes from 0.46 to 0.90, such that providing an accurate scale factor between both values is difficult. Furthermore, the ΔE00/ΔE∗ab ratio increases with the increase in ΔL∗ and the decrease in Δb∗. The pairs for which the ratio is highest are those in which the difference in color is mainly due to changes in lightness, whereas the pairs for which the ratio is the smallest are those in which the difference in color is mainly due to changes on the blue-yellow axis.

Clinical study on natural gingival color

The aims of the study were: to describe the gingival color surrounding the upper incisors in three sites in the keratinized gingiva, analyzing the effect of possible factors which modulate (socio-demographic and behavioral) intersubject variability; to study whether the gingiva color is the same in all three locations and to describe intrasubject color differences in the keratinized gingiva band. Using the CIELAB color system, three reference areas (free gingival margin, keratinized gingival body, and birth or upper part of the keratinized gingiva) were studied in 259 individuals, as well as the related socio-demographic factors, oral habits and the chronic intake of medication. Shadepilot™ spectrophotometer was used. Descriptive and inferential statistical analysis was performed. There are statistically significant differences between males and females for coordinates L* and a* in the middle and free gingival margin. For the b* coordinate, there are differences between males and females in the three locations studied (p < 0.05). The minimum and maximum coordinates in which the CIELAB natural gingival space is delimited are L* minima 28.3, L* maximum 65.4, a* minimum 11.1, a* maximum 37.2, b* minimum 6.9, and b* maximum 25.2*. Age, smoking, and the chronic intake of medication had no significant effect on gum color. There are perceptible color differences within the keratinized gingiva band. These chromatic differences must be taken into account if the prosthetic characterization of gingival tissue is to be considered acceptable. There are significant differences between the color coordinates of the three sites studied in the keratinized gingiva of men and women.

Comparison of the CIELab and CIEDE 2000 Color Difference Formulas on Gingival Color Space: Color-Difference Formulas on Gingiva Color Space

PURPOSE To evaluate the CIELAB and CIEDE2000 formulas to determine which better reflects the difference in color perception, and whether there are gender differences in color perception. MATERIALS AND METHODS Fifty-six participants grouped 21 different-colored disks made of pink porcelain (color range: from 36.8 to 63.9 for L* , from 16.4 to 35.2 for a* , from 6.2 to 21.6 for b* ), in which the only requirement was that each group be comprised of disks with chromatically indistinguishable colors. Each participant was free to choose the number and composition of the groups. Using the results obtained, a dissimilarity matrix was generated, and nonmetric multidimensional scaling (MDS) was applied to obtain the coordinates of the disks within a Euclidean space. RESULTS The linear correlation coefficient between the interpoint distances of the MDS configuration (MDS_total sample) and the color differences with the CIELAB formula (ΔE*ab ) was 0.719 (p < 0.001), whereas with the CIEDE2000 formula (ΔE00), it was 0.726 (p < 0.001). For the configuration obtained in the womens group (MDS_w), the correlations between the interpoint distances and the CIELab color difference was 0.720 (p < 0.001), and the CIEDE2000 color differences was 0.730 (p < 0.001). For the configuration obtained in the mens group (MDS_m), the perceived color differences, calculated with both formulas, were less sensitive, 0.670 for CIELab formula and 0.677 for CIEDE2000 formula. CONCLUSIONS Considering the limitations of this study, the data analyzed indicated that the CIEDE2000 formula reflected the color differences perceived by the human eye similar to the CIELAB formula (ΔE*ab ). Also, women were confirmed to be more sensitive than men to differences in color.

Alternativa al análisis canónico de correspondencias basada en los métodos Biblot

Los estudios de comunidades ecologicas implican, con frecuencia, el inferir las relaciones existentes entre las especies y las variables ambientales a partir de los datos de composicion de la comunidad y de medidas asociadas al habitat. Es un hecho que la composiciun de la comunidad cambia a lo largo de gradientes ambientales y que las especies tienen dentro de estos un optimo ambiental y un rango concretos. Sin embargo los gradientes ambientales no son directamente observables pero son imprescindibles para poder explicar los reemplazamientos tanto espaciales como temporales de unas especies por otras, cuando varian las condiciones del medio. En este trabajo, se propone el HJ-B1PLOT como alternativa al Analisis Canonico de Correspondencias (CCA) (ter BRAAK, 1986) siendo tambien una tecnica de analisis multivariante directo del gradiente. El HJ-BIPLOT proporciona mejores resultados que el CCA ya que obtiene la misma calidad de representaciun para filas y columnas. Se realiza una aplicacion a datos reales, y se determinan los optimos de las especies respecto de las variables analizadas y el ranking de las mismas en relacion al gradiente para dichas variables.