- Email: [email protected]
A new approach to the evaluation of whitening effect of a cosmetic using computer analysis of video-captured image
Journal of Dermatological Science 29 (2002) 10 – 18 www.elsevier.com/locate/jdermsci
A new approach to the evaluation of whitening effect of a cosmetic using computer analysis of video-captured image Akira Kawada a,*, Hiroko Kameyama a, Mutsuyo Asai a, Hatsuki Shiraishi a, Yoshinori Aragane a, Tadashi Tezuka a, Kouji Iwakiri b a
Department of Dermatology, Kinki Uni6ersity School of Medicine, Ohno-Higashi 377 -2, Osaka-Sayama City, Osaka 589 -8511, Japan b Inforward Inc., Shibuya-ku, Tokyo 151 -0053, Japan Received 10 October 2001; received in revised form 5 December 2001; accepted 6 December 2001
Abstract Whitening effects of cosmetics have been evaluated with change of skin reflectance of the colorimeter. However, skin color implies Hue (H) and Saturation (S) as well as skin reflectance (Value). We have developed a new evaluation method of change of skin color using computer analysis of the video-captured digital image. We have also investigated whitening effects of a new whitening cosmetic essence, ‘Concentre anti-tache nuit’ (CAN; Parfums Christian Dior, France), which contained 3% magnesium l-ascorbyl 2-phosphate, on skin color of the face in 15 healthy Japanese females. Measurement was performed after 6 weeks application. Whitening effects were also evaluated with the combination of observation and photographs. CAN showed significant improvement in Saturation of the forehead and left cheek and Value of the forehead. CAN showed greater whitening effects both on the forehead and left cheek in eight (53%) of 15 subjects with the combination of observation and photographs. These results indicated that CAN may have a whitening effect on skin of the face after 6 weeks application. This method may be useful to evaluate the whitening effect of cosmetics and color change of cutaneous disorders. © 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Video-captured digital image; Skin color; HSV system; Whitening effect
1. Introduction A preference for a fair skin is a common recent trend amongst women in Asian countries. In Japan, sales of whitening cosmetics are increasing. Whitening cosmetics are expected to improve * Corresponding author. Tel.: +81-72-366-0221; fax: + 8172-368-2120. E-mail address: [email protected] (A. Kawada).
dullness as well as lightness. Skin reflectance (lightness) has been used as a measure of the evaluation of whitening effects against chloasma [1,2] and UV-induced delayed tanning [3,4]. Complexion is comprised of lightness, dullness, and color coordinates. Whitening cosmetics may affect melanin that absorbs throughout spectrum of the visible light. Both reflectance and color coordinates should be considered in the evaluation of whitening cosmetics.
0923-1811/02/$ - see front matter © 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 9 2 3 - 1 8 1 1 ( 0 1 ) 0 0 1 7 3 - 6
A. Kawada et al. / Journal of Dermatological Science 29 (2002) 10–18
Measurement of skin color can be made by using two different principles: spectrophotometric, using either a broad band (scanning) or selected wavelengths in the visible range with measurement of absorbance and reflectance; and tristimulus (blue, red, green) analysis of light reflected from skin structures [5]. For quantification of skin color and erythema, two types of instruments, a narrow-band spectrophotometer (DermaSpectrometer, Cortex Technology, Hadsund, Denmark) and a tristimulus colorimeter (Minolta Chroma Meter, Osaka, Japan) are commonly used [5,6]. CIELAB color space system has been widely accepted to describe a color [5,7], although many different systems including Munsell, L*C*h, Hunter’s Lab, and XYZ (Yxy) color systems have been developed to evaluate color as an objective measure of the human visual perception of colors. L*, a*, and b* are defined as brightness (white –black), red– green chromacity coordinate, and blue– yellow chromacity coordinate in CIELAB color space. This CIELAB color space is also used to evaluate the activity of blanching products [8,9]. However, CIELAB in the tristimulus colorimeter may not detect the slight difference in skin color when using whitening cosmetics, because these cosmetics usually do not have strong bleaching effects such as hydroquinone creams. In the dermatological field, the combination of videomicroscope and computer has been examined to quantify erythema, pigmentation, and color of skin lesions [10,11]. They gained the brightness for each band of red, green, and blue (RGB) and calculated the integrated reflectance and absorbance index. Their system expresses every color with the conversion of the integrated reflectance values into quasi-Munsell color space. In this study, we have developed a new measurement system of skin color using the computer analysis of the video-captured digital image. In our system, skin color was firstly identified as RGB data and was converted to HSV (Hue (H), Saturation (S), Value (V)) data with Smith’s theory [12]. We have recently demonstrated whitening effects of a new whitening cosmetic essence, ‘Concentre anti-tache nuit’ (CAN; Parfums Christian Dior, France), against UV-induced delayed
11
tanning [13]. Then, we report whitening effects of CAN on facial skin color using our new HSV system as a measure.
2. Subjects and methods
2.1. Subjects Fifteen healthy Japanese females aged 18–51 years, who were receiving no medication, participated in this study that was carried out between September and November of 2000. Skin type of all the subjects was classified as J-II in Japanese skin type classification [14]. Subjects who had extensive pigmentary disorders, e.g. cholasma, ephelides, and multiple solar lentigines on the face were excluded from the study. CAN was applied for all the face once a day at night for 6 weeks. The forehead and left cheek of the face were used for the evaluation. During the study, recreational exposure to sunlight without sunscreens was inhibited. Informed consent was obtained.
2.2. Testing substance A beauty lotion, ‘CAN’ (Parfums Christian Dior), was used. CAN contained 3% magnesium l-ascorbyl 2-phosphate (APC-3), vitamin A, and antisense compound (DNA Na). This preparation has been proved to have whitening effects against UV-induced delayed tanning as described previously [13].
2.3. System of computer analysis Inforward Imaging System VI-10 (Inforward Inc., Tokyo) was used for the evaluation. The skin images were taken with a 1/4-inch charge coupled device (CCD) camera (270 000 pixels). The magnification was 40-fold. Capturing device was a CCD image sensor that kept highly uniformed brightness. This device was connected to the Windows-based Personal Computer via a digital video unit and an image acquisition PC card. Skin images captured with this digital system were displayed as image patterns of histogram using the software analysis. In each color parameter
12
A. Kawada et al. / Journal of Dermatological Science 29 (2002) 10–18
(Red, Green, Blue, Hue, Saturation, and Value, HSV), 8-bit (28 = 256) data were obtained from CCD-captured images. In a scale of each color (0–255°), the software calculates number of pixels
at each degree and then expresses a histogram. The histograms were demonstrated either in RGB or HSV system (Fig. 1). HSV values were converted from RGB data using Smith’s HSV color
Fig. 1. A video-captured image and histograms of Hue, Saturation and Value on the forehead of a representative case.
A. Kawada et al. / Journal of Dermatological Science 29 (2002) 10–18
13
Table 1 Hue, Saturation and Value of the forehead skin with ‘CAN’-application before and after 6 weeks Number
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Mean S.D.
Hue
Saturation
Value
Before
6 weeks
Before
6 weeks
Before
6 weeks
17.1 19.4 19.2 16.8 18.0 17.3 18.7 15.9 15.6 21.6 16.6 18.7 19.9 21.1 21.1 18.5 1.9
19.4 18.3 19.0 20.1 22.0 18.5 19.7 16.4 19.0 20.4 19.7 19.4 19.0 21.3 20.1 19.5 1.3
82.8 78.9 76.4 92.6 64.0 73.4 73.4 84.9 77.7 68.7 77.7 87.9 87.0 84.1 62.3 78.1 8.8
86.2 80.6 85.3 81.5 82.8 77.2 87.9 81.5 96.9 84.5 75.5 97.7 101.1 78.5 68.7 84.4* 8.8
163.8 160.9 163.4 170.2 157.0 163.0 163.4 165.5 168.1 157.9 162.6 166.0 164.7 163.0 159.6 163.3 3.5
165.1 164.7 165.1 163.0 164.7 164.3 166.4 168.1 169.0 165.5 161.7 169.4 167.7 163.4 160.9 165.3* 2.5
*, PB0.05 compared with before application.
models [12]. From these histograms, the representative data were calculated as figures with this software. Calibration was performed before and after the measurement using the adjustment of white balance of CCD image sensor. Using the software, small pigmentary spots and hair were removed from the captured image and color data were evaluated.
2.4. E6aluation of whiteness of face skin Firstly, captured skin images were analyzed with the system described above. We decided the each center point of forehead and left cheek. On the forehead, two lateral points with the margin of 2.5 cm from the center point were fixed using a measure. On the left cheek, two lateral, upper, and lower points with the margin of 2.5 cm from the center point were fixed using a measure. In three (center, left, and right) and five (center, left, right, upper, and lower) fixed areas of the forehead and left cheek of each subject, the average of Hue (°), Saturation (%), and Value (%) were calculated before application and after 6 weeks.
For the second, evaluation was also conducted by the combination of observation and photographs with three investigators (Akira Kawada, Hatsuki Shirashi, and Mutsuyo Asai.). Whiteness of face skin was scored from one to five; one, very white; two, white; three, usual complexion as Japanese females; four, dark; and five, very dark. Scores in the forehead and left cheek were recorded before application and after 6 weeks. Additionally, L*, a*, and b* values in three and five fixed areas of the forehead and left cheek were measured using Minolta Chroma Meter CR-221 (Minolta). Three fixed areas of the untreated left arm were set as the control of CAN-treated area and the color data of Hue, Saturation, Value, L*, a*, and b* before and after the test.
3. Results HSV data of the forehead and left cheek were shown in Tables 1 and 2, respectively. Using t-test, significant differences of means of S both in the forehead and left cheek and Value in the
A. Kawada et al. / Journal of Dermatological Science 29 (2002) 10–18
14
forehead were shown between pre- and post-application (Tables 1 and 2). The representative data of the captured images and HSV histograms were shown in Figs. 2 and 3. In each histogram, the shift of peaks from left to right corresponds to change of Hue, Satiration, or Value after 6 weeks application. After 6 weeks application, the improvement of whiteness by the combination of observation and photographs was found in eight (57%) of 15 subjects both on the forehead and left cheek with no change in seven subjects and worsening in no subjects. This result by observation and photographs supports the results of Saturation and Value. CIELAB date of the forehead and left cheek were shown in Tables 3 and 4, respectively. L*, a*, and b* values both in the forehead and left cheek did not show significant differences between pre- and post-application. Values of c* calculated from (a*2 +b*2)0.5 also did not show the significant differences between pre- and postapplication. In the control area without treatment of CAN, no significant differences of Hue, Saturation, Value, L*, a*, or b* data were found between pre- and post-treatments.
4. Discussion Whitening cosmetics in Asian countries should have minimal bleaching effects on existed melanin in the skin, since they are used for the improvement of constitutional complexion of the face. Therefore, the instruments that detect slight changes in skin color are needed for the evaluation of whitening cosmetics. In our study, CIELAB color spaces of the tristimulus colorimeter used did not demonstrate the significant difference in skin color after application of CAN, whereas Saturation and Value in HSV data significantly improved with use of CAN. This difference may have been due to the difference of sensitivity in the instruments. HSV data acquired from CCD camera in our system may be useful for the measurement of constitutional skin color and minimal changes of skin color. RGB space is a physicochemical expression of colors on RGB monitors that does not reflect human visual perception. Munsell color space implies human visual perception and expresses color as Hue, Value, and Chroma. However, Munsell
Table 2 Hue, Saturation and Value of the skin of left cheek with ‘CAN’-application before and after 6 weeks Number
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Mean S.D.
Hue (°)
Saturation
Value (%)
Before
6 weeks
Before
6 weeks
Before
6 weeks
14.2 18.6 16.9 14.1 15.6 14.2 17.1 15.9 17.2 19.3 14.9 17.4 16.1 17.5 17.2 16.4 1.6
16.4 18.2 18.8 17.1 19.2 14.5 15.1 13.2 16.9 17.2 15.1 13.5 17.2 16.4 17.4 16.4 1.8
76.3 73.5 79.9 84.7 72.4 73.7 82.2 75.0 81.4 90.4 72.7 82.2 101.4 99.3 88.1 82.2 9.2
73.0 76.5 78.8 81.2 82.4 65.8 86.0 82.9 86.3 87.3 79.9 89.3 104.2 107.5 89.3 84.7* 10.7
166.5 160.3 166.1 166.5 163.6 165.4 164.6 165.1 166.1 167.2 161.8 166.4 171.8 169.7 167.7 165.9 2.8
163.3 163.6 164.9 162.3 165.4 161.8 165.4 170.8 166.7 166.1 165.9 168.4 171.5 174.1 166.7 166.5 3.5
*, PB0.05 compared with before application.
A. Kawada et al. / Journal of Dermatological Science 29 (2002) 10–18
15
Fig. 2. Video-captured images and histograms of Hue, Saturation and Value on the forehead skin of case number 9 with ‘CAN’-application before and after 6 weeks in a representative case.
color space is ordinarily inappropriate for digital image processing. Therefore, Hue, Saturation, and Intensity (I) are used as digital parameters for
the visual perception [15]. The HSI conversion from RGB data has been developed to match the monitor’s RGB data with Munsell color space
16
A. Kawada et al. / Journal of Dermatological Science 29 (2002) 10–18
[15]. In HSI conversion systems, HSV (Hue, Saturation, and V) [12] and HSL (Hue, Saturation, and Luminosity) [16] color models are commonly
used. We have chosen HSV system because HSV system can more precisely detect difference in color [17].
Fig. 3. Video-captured images and histograms of Hue, Saturation and Value on the skin of left cheek of case number 5 with ‘CAN’-application before and after 6 weeks in a representative case.
A. Kawada et al. / Journal of Dermatological Science 29 (2002) 10–18
17
Table 3 L*, a*, and b* of the forehead skin with ‘CAN’-application before and after 6 weeks Number
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Mean S.D.
L*
a*
b*
Before
6 weeks
Before
6 weeks
Before
6 weeks
55.5 57.2 56.3 54.6 58.9 54.0 56.6 52.2 53.4 55.1 57.7 54.2 55.5 52.8 58.3 55.5 1.9
52.9 55.3 55.8 55.0 56.3 56.5 57.4 57.8 53.4 55.3 55.5 54.3 54.9 55.5 57.9 55.6 1.4
6.1 5.9 5.3 5.7 3.5 9.4 6.1 12.7 8.1 4.1 4.5 8.0 5.6 6.9 5.1 6.5 2.3
8.4 9.1 4.8 6.8 7.1 6.3 4.8 6.4 7.5 4.1 6.9 6.8 4.9 3.8 3.6 6.1 1.6
11.1 12.7 12.2 11.2 10.5 9.6 10.8 11.0 11.5 10.9 10.4 14.4 14.5 11.3 11.4 11.5 1.3
11.7 12.0 13.0 10.9 11.1 10.7 11.9 11.6 11.9 10.8 10.5 15.3 15.1 12.5 11.4 12.0 1.4
Table 4 L*, a*, and b* of the skin of left cheek with ‘CAN’-application before and after 6 weeks Number
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Mean S.D.
L*
a*
b*
Before
6 weeks
Before
6 weeks
Before
6 weeks
60.1 63.6 58.0 58.9 59.3 58.8 59.4 56.5 53.9 55.0 55.7 57.9 56.8 52.2 61.3 57.8 2.8
61.1 62.3 60.9 60.9 59.6 58.5 56.0 59.0 54.2 53.3 57.0 59.5 53.3 53.2 62.2 58.1 3.2
3.9 1.7 6.7 4.6 4.8 5.2 2.7 8.9 6.8 4.0 5.6 4.5 5.8 7.3 4.4 5.1 1.7
3.1 2.4 3.7 3.4 4.1 4.4 7.5 7.5 6.1 7.4 5.5 4.2 5.9 6.8 5.1 5.1 1.6
9.6 9.6 9.3 8.7 8.6 5.7 11.5 9.4 11.4 9.4 10.0 12.5 12.2 10.9 10.6 10.0 1.6
10.8 10.3 10.0 8.4 11.6 7.0 11.1 9.0 11.9 9.2 9.9 12.5 14.7 12.5 10.0 10.6 1.8
The measurement of HSV data showed significant improvement in Saturation of the forehead and left cheek and Value of the forehead. In eight
(57%) of 15 Japanese female subjects CAN had whitening effects on the skin of forehead and left cheek with the measurement of the combination
18
A. Kawada et al. / Journal of Dermatological Science 29 (2002) 10–18
of observation and photographs. CAN, therefore, may have whitening effects on skin of the face in the subjects after 6 weeks application. ‘Kusumi’ is a Japanese term that means dullness in color. ‘Kusumi’ collects very much attention from Japanese females as well as whiteness. However, there have been no measurement methods of ‘Kusumi’. Our new method has measured the change in Saturation, suggesting that our method may be used in evaluation of effects of cosmetics on Kusumi.
References [1] Haramoto I, Mizoguchi M. Clinical evaluation of licorice extracts on chloasma. Nishinihon J Dermatol 1995;57:601 – 8 in Japanese. [2] Kita Y, Sugai T. Effect of bleach agents on chloasma. Hifu 1992;34:142 – 6 in Japanese. [3] Sugai T. Recent bleach products seeing from the standpoint of dermatology and scientific evaluation of bleaching effects. Fragrance J 1990;18:32 – 8 in Japanese. [4] Morimoto K, Kawada A, Hiruma M, Ishibashi A. An evaluation of whitening effect of Serum Clarifiant against UVB-induced delayed tanning. Acta Dermatol (Kyoto) 1994;89:675 – 8. [5] Fullerton A, Fischer T, Lahti A, Wilhelm KP, Takiwaki H, Serup J. Guidelines for measurement of skin colour and erythema. A report from the Standardization Group of the European society of contact dermatitis. Contact Dermatitis 1996;35:1 –10. [6] Takiwaki H, Overgaard L, Serup J. Comparison of narrow-band reflectance spectrophotometric and tristimulus colorimetric measurements of skin color. Skin Pharmacol
1994;7:217 – 25. [7] Weatherall IL, Coombs BD. Skin color measurements in terms of CIELAB color space values. J Invest Dermatol 1992;99:468 – 73. [8] Duteil L, Ortonne JP. Colorimetric assessment of the effects of azelaic acid on light-induced skin pigmentation. Photodermatol Photoimmunol Photomed 1992;9:67 – 71. [9] Andreassi L, Flori L. Practical application of cutaneous colorimetry. Clin Dermatol 1995;13:369 – 73. [10] Takiwaki H, Shirai S, Kanno Y, Watanabe Y, Arase S. Quantification of erythema and pigmentation using a videomicroscope and a computer. Br J Dermatol 1994;131:85 – 92. [11] Takiwaki H, Shirai S, Watanabe Y, Nakagawa K, Arase S. A rudimentary system for automatic discrimination among basic skin lesions on the basis of color analysis of video images. J Am Acad Dermatol 1995;32:600 – 4. [12] Smith AR. Color gamut transformation pairs. Comput Graph (ACM) 1978;12:12 – 9. [13] Kawada A, Shiraishi H, Asai M, Aragane Y, Tezuka T. An evaluation of whitening effect of Concentre Antitache Nuit against solar simulated radiation-induced delayed tanning. Skin Res 2001;43:5 – 9. [14] Satoh Y, Kawada A. Action spectrum for melanin pigmentation to ultraviolet light, and Japanese skin typing. In: Fitzpatrick TB, Wick MM, Toda K, editors. Brown Melanoderma. Biology and Disease of Epidermal Pigmentation. Tokyo: University of Tokyo Press, 1986:87 – 95. [15] Buchanan MD, Pendergrass R. Digital image processing: can intensity, hue, and saturation replace red, green, and blue. Electro-Optical Syst Des 1980;12:29 – 36. [16] Joblore GH, Greenberg D. Color spaces for computer graphics. Comput Graph (ACM) 1978;12:20 – 7. [17] Fukue K, Shimoda A, Sakata T. Evaluation of various color models of HIS, in: Proceedings of 1986 Autumn Meeting of Japan Society of Photogrammetry and Remote Sensing, 1986:63 – 68.
A new approach to the evaluation of whitening effect of a cosmetic using computer analysis of video-captured image Akira Kawada a,*, Hiroko Kameyama a, Mutsuyo Asai a, Hatsuki Shiraishi a, Yoshinori Aragane a, Tadashi Tezuka a, Kouji Iwakiri b a
Department of Dermatology, Kinki Uni6ersity School of Medicine, Ohno-Higashi 377 -2, Osaka-Sayama City, Osaka 589 -8511, Japan b Inforward Inc., Shibuya-ku, Tokyo 151 -0053, Japan Received 10 October 2001; received in revised form 5 December 2001; accepted 6 December 2001
Abstract Whitening effects of cosmetics have been evaluated with change of skin reflectance of the colorimeter. However, skin color implies Hue (H) and Saturation (S) as well as skin reflectance (Value). We have developed a new evaluation method of change of skin color using computer analysis of the video-captured digital image. We have also investigated whitening effects of a new whitening cosmetic essence, ‘Concentre anti-tache nuit’ (CAN; Parfums Christian Dior, France), which contained 3% magnesium l-ascorbyl 2-phosphate, on skin color of the face in 15 healthy Japanese females. Measurement was performed after 6 weeks application. Whitening effects were also evaluated with the combination of observation and photographs. CAN showed significant improvement in Saturation of the forehead and left cheek and Value of the forehead. CAN showed greater whitening effects both on the forehead and left cheek in eight (53%) of 15 subjects with the combination of observation and photographs. These results indicated that CAN may have a whitening effect on skin of the face after 6 weeks application. This method may be useful to evaluate the whitening effect of cosmetics and color change of cutaneous disorders. © 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Video-captured digital image; Skin color; HSV system; Whitening effect
1. Introduction A preference for a fair skin is a common recent trend amongst women in Asian countries. In Japan, sales of whitening cosmetics are increasing. Whitening cosmetics are expected to improve * Corresponding author. Tel.: +81-72-366-0221; fax: + 8172-368-2120. E-mail address: [email protected] (A. Kawada).
dullness as well as lightness. Skin reflectance (lightness) has been used as a measure of the evaluation of whitening effects against chloasma [1,2] and UV-induced delayed tanning [3,4]. Complexion is comprised of lightness, dullness, and color coordinates. Whitening cosmetics may affect melanin that absorbs throughout spectrum of the visible light. Both reflectance and color coordinates should be considered in the evaluation of whitening cosmetics.
0923-1811/02/$ - see front matter © 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 9 2 3 - 1 8 1 1 ( 0 1 ) 0 0 1 7 3 - 6
A. Kawada et al. / Journal of Dermatological Science 29 (2002) 10–18
Measurement of skin color can be made by using two different principles: spectrophotometric, using either a broad band (scanning) or selected wavelengths in the visible range with measurement of absorbance and reflectance; and tristimulus (blue, red, green) analysis of light reflected from skin structures [5]. For quantification of skin color and erythema, two types of instruments, a narrow-band spectrophotometer (DermaSpectrometer, Cortex Technology, Hadsund, Denmark) and a tristimulus colorimeter (Minolta Chroma Meter, Osaka, Japan) are commonly used [5,6]. CIELAB color space system has been widely accepted to describe a color [5,7], although many different systems including Munsell, L*C*h, Hunter’s Lab, and XYZ (Yxy) color systems have been developed to evaluate color as an objective measure of the human visual perception of colors. L*, a*, and b* are defined as brightness (white –black), red– green chromacity coordinate, and blue– yellow chromacity coordinate in CIELAB color space. This CIELAB color space is also used to evaluate the activity of blanching products [8,9]. However, CIELAB in the tristimulus colorimeter may not detect the slight difference in skin color when using whitening cosmetics, because these cosmetics usually do not have strong bleaching effects such as hydroquinone creams. In the dermatological field, the combination of videomicroscope and computer has been examined to quantify erythema, pigmentation, and color of skin lesions [10,11]. They gained the brightness for each band of red, green, and blue (RGB) and calculated the integrated reflectance and absorbance index. Their system expresses every color with the conversion of the integrated reflectance values into quasi-Munsell color space. In this study, we have developed a new measurement system of skin color using the computer analysis of the video-captured digital image. In our system, skin color was firstly identified as RGB data and was converted to HSV (Hue (H), Saturation (S), Value (V)) data with Smith’s theory [12]. We have recently demonstrated whitening effects of a new whitening cosmetic essence, ‘Concentre anti-tache nuit’ (CAN; Parfums Christian Dior, France), against UV-induced delayed
11
tanning [13]. Then, we report whitening effects of CAN on facial skin color using our new HSV system as a measure.
2. Subjects and methods
2.1. Subjects Fifteen healthy Japanese females aged 18–51 years, who were receiving no medication, participated in this study that was carried out between September and November of 2000. Skin type of all the subjects was classified as J-II in Japanese skin type classification [14]. Subjects who had extensive pigmentary disorders, e.g. cholasma, ephelides, and multiple solar lentigines on the face were excluded from the study. CAN was applied for all the face once a day at night for 6 weeks. The forehead and left cheek of the face were used for the evaluation. During the study, recreational exposure to sunlight without sunscreens was inhibited. Informed consent was obtained.
2.2. Testing substance A beauty lotion, ‘CAN’ (Parfums Christian Dior), was used. CAN contained 3% magnesium l-ascorbyl 2-phosphate (APC-3), vitamin A, and antisense compound (DNA Na). This preparation has been proved to have whitening effects against UV-induced delayed tanning as described previously [13].
2.3. System of computer analysis Inforward Imaging System VI-10 (Inforward Inc., Tokyo) was used for the evaluation. The skin images were taken with a 1/4-inch charge coupled device (CCD) camera (270 000 pixels). The magnification was 40-fold. Capturing device was a CCD image sensor that kept highly uniformed brightness. This device was connected to the Windows-based Personal Computer via a digital video unit and an image acquisition PC card. Skin images captured with this digital system were displayed as image patterns of histogram using the software analysis. In each color parameter
12
A. Kawada et al. / Journal of Dermatological Science 29 (2002) 10–18
(Red, Green, Blue, Hue, Saturation, and Value, HSV), 8-bit (28 = 256) data were obtained from CCD-captured images. In a scale of each color (0–255°), the software calculates number of pixels
at each degree and then expresses a histogram. The histograms were demonstrated either in RGB or HSV system (Fig. 1). HSV values were converted from RGB data using Smith’s HSV color
Fig. 1. A video-captured image and histograms of Hue, Saturation and Value on the forehead of a representative case.
A. Kawada et al. / Journal of Dermatological Science 29 (2002) 10–18
13
Table 1 Hue, Saturation and Value of the forehead skin with ‘CAN’-application before and after 6 weeks Number
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Mean S.D.
Hue
Saturation
Value
Before
6 weeks
Before
6 weeks
Before
6 weeks
17.1 19.4 19.2 16.8 18.0 17.3 18.7 15.9 15.6 21.6 16.6 18.7 19.9 21.1 21.1 18.5 1.9
19.4 18.3 19.0 20.1 22.0 18.5 19.7 16.4 19.0 20.4 19.7 19.4 19.0 21.3 20.1 19.5 1.3
82.8 78.9 76.4 92.6 64.0 73.4 73.4 84.9 77.7 68.7 77.7 87.9 87.0 84.1 62.3 78.1 8.8
86.2 80.6 85.3 81.5 82.8 77.2 87.9 81.5 96.9 84.5 75.5 97.7 101.1 78.5 68.7 84.4* 8.8
163.8 160.9 163.4 170.2 157.0 163.0 163.4 165.5 168.1 157.9 162.6 166.0 164.7 163.0 159.6 163.3 3.5
165.1 164.7 165.1 163.0 164.7 164.3 166.4 168.1 169.0 165.5 161.7 169.4 167.7 163.4 160.9 165.3* 2.5
*, PB0.05 compared with before application.
models [12]. From these histograms, the representative data were calculated as figures with this software. Calibration was performed before and after the measurement using the adjustment of white balance of CCD image sensor. Using the software, small pigmentary spots and hair were removed from the captured image and color data were evaluated.
2.4. E6aluation of whiteness of face skin Firstly, captured skin images were analyzed with the system described above. We decided the each center point of forehead and left cheek. On the forehead, two lateral points with the margin of 2.5 cm from the center point were fixed using a measure. On the left cheek, two lateral, upper, and lower points with the margin of 2.5 cm from the center point were fixed using a measure. In three (center, left, and right) and five (center, left, right, upper, and lower) fixed areas of the forehead and left cheek of each subject, the average of Hue (°), Saturation (%), and Value (%) were calculated before application and after 6 weeks.
For the second, evaluation was also conducted by the combination of observation and photographs with three investigators (Akira Kawada, Hatsuki Shirashi, and Mutsuyo Asai.). Whiteness of face skin was scored from one to five; one, very white; two, white; three, usual complexion as Japanese females; four, dark; and five, very dark. Scores in the forehead and left cheek were recorded before application and after 6 weeks. Additionally, L*, a*, and b* values in three and five fixed areas of the forehead and left cheek were measured using Minolta Chroma Meter CR-221 (Minolta). Three fixed areas of the untreated left arm were set as the control of CAN-treated area and the color data of Hue, Saturation, Value, L*, a*, and b* before and after the test.
3. Results HSV data of the forehead and left cheek were shown in Tables 1 and 2, respectively. Using t-test, significant differences of means of S both in the forehead and left cheek and Value in the
A. Kawada et al. / Journal of Dermatological Science 29 (2002) 10–18
14
forehead were shown between pre- and post-application (Tables 1 and 2). The representative data of the captured images and HSV histograms were shown in Figs. 2 and 3. In each histogram, the shift of peaks from left to right corresponds to change of Hue, Satiration, or Value after 6 weeks application. After 6 weeks application, the improvement of whiteness by the combination of observation and photographs was found in eight (57%) of 15 subjects both on the forehead and left cheek with no change in seven subjects and worsening in no subjects. This result by observation and photographs supports the results of Saturation and Value. CIELAB date of the forehead and left cheek were shown in Tables 3 and 4, respectively. L*, a*, and b* values both in the forehead and left cheek did not show significant differences between pre- and post-application. Values of c* calculated from (a*2 +b*2)0.5 also did not show the significant differences between pre- and postapplication. In the control area without treatment of CAN, no significant differences of Hue, Saturation, Value, L*, a*, or b* data were found between pre- and post-treatments.
4. Discussion Whitening cosmetics in Asian countries should have minimal bleaching effects on existed melanin in the skin, since they are used for the improvement of constitutional complexion of the face. Therefore, the instruments that detect slight changes in skin color are needed for the evaluation of whitening cosmetics. In our study, CIELAB color spaces of the tristimulus colorimeter used did not demonstrate the significant difference in skin color after application of CAN, whereas Saturation and Value in HSV data significantly improved with use of CAN. This difference may have been due to the difference of sensitivity in the instruments. HSV data acquired from CCD camera in our system may be useful for the measurement of constitutional skin color and minimal changes of skin color. RGB space is a physicochemical expression of colors on RGB monitors that does not reflect human visual perception. Munsell color space implies human visual perception and expresses color as Hue, Value, and Chroma. However, Munsell
Table 2 Hue, Saturation and Value of the skin of left cheek with ‘CAN’-application before and after 6 weeks Number
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Mean S.D.
Hue (°)
Saturation
Value (%)
Before
6 weeks
Before
6 weeks
Before
6 weeks
14.2 18.6 16.9 14.1 15.6 14.2 17.1 15.9 17.2 19.3 14.9 17.4 16.1 17.5 17.2 16.4 1.6
16.4 18.2 18.8 17.1 19.2 14.5 15.1 13.2 16.9 17.2 15.1 13.5 17.2 16.4 17.4 16.4 1.8
76.3 73.5 79.9 84.7 72.4 73.7 82.2 75.0 81.4 90.4 72.7 82.2 101.4 99.3 88.1 82.2 9.2
73.0 76.5 78.8 81.2 82.4 65.8 86.0 82.9 86.3 87.3 79.9 89.3 104.2 107.5 89.3 84.7* 10.7
166.5 160.3 166.1 166.5 163.6 165.4 164.6 165.1 166.1 167.2 161.8 166.4 171.8 169.7 167.7 165.9 2.8
163.3 163.6 164.9 162.3 165.4 161.8 165.4 170.8 166.7 166.1 165.9 168.4 171.5 174.1 166.7 166.5 3.5
*, PB0.05 compared with before application.
A. Kawada et al. / Journal of Dermatological Science 29 (2002) 10–18
15
Fig. 2. Video-captured images and histograms of Hue, Saturation and Value on the forehead skin of case number 9 with ‘CAN’-application before and after 6 weeks in a representative case.
color space is ordinarily inappropriate for digital image processing. Therefore, Hue, Saturation, and Intensity (I) are used as digital parameters for
the visual perception [15]. The HSI conversion from RGB data has been developed to match the monitor’s RGB data with Munsell color space
16
A. Kawada et al. / Journal of Dermatological Science 29 (2002) 10–18
[15]. In HSI conversion systems, HSV (Hue, Saturation, and V) [12] and HSL (Hue, Saturation, and Luminosity) [16] color models are commonly
used. We have chosen HSV system because HSV system can more precisely detect difference in color [17].
Fig. 3. Video-captured images and histograms of Hue, Saturation and Value on the skin of left cheek of case number 5 with ‘CAN’-application before and after 6 weeks in a representative case.
A. Kawada et al. / Journal of Dermatological Science 29 (2002) 10–18
17
Table 3 L*, a*, and b* of the forehead skin with ‘CAN’-application before and after 6 weeks Number
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Mean S.D.
L*
a*
b*
Before
6 weeks
Before
6 weeks
Before
6 weeks
55.5 57.2 56.3 54.6 58.9 54.0 56.6 52.2 53.4 55.1 57.7 54.2 55.5 52.8 58.3 55.5 1.9
52.9 55.3 55.8 55.0 56.3 56.5 57.4 57.8 53.4 55.3 55.5 54.3 54.9 55.5 57.9 55.6 1.4
6.1 5.9 5.3 5.7 3.5 9.4 6.1 12.7 8.1 4.1 4.5 8.0 5.6 6.9 5.1 6.5 2.3
8.4 9.1 4.8 6.8 7.1 6.3 4.8 6.4 7.5 4.1 6.9 6.8 4.9 3.8 3.6 6.1 1.6
11.1 12.7 12.2 11.2 10.5 9.6 10.8 11.0 11.5 10.9 10.4 14.4 14.5 11.3 11.4 11.5 1.3
11.7 12.0 13.0 10.9 11.1 10.7 11.9 11.6 11.9 10.8 10.5 15.3 15.1 12.5 11.4 12.0 1.4
Table 4 L*, a*, and b* of the skin of left cheek with ‘CAN’-application before and after 6 weeks Number
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Mean S.D.
L*
a*
b*
Before
6 weeks
Before
6 weeks
Before
6 weeks
60.1 63.6 58.0 58.9 59.3 58.8 59.4 56.5 53.9 55.0 55.7 57.9 56.8 52.2 61.3 57.8 2.8
61.1 62.3 60.9 60.9 59.6 58.5 56.0 59.0 54.2 53.3 57.0 59.5 53.3 53.2 62.2 58.1 3.2
3.9 1.7 6.7 4.6 4.8 5.2 2.7 8.9 6.8 4.0 5.6 4.5 5.8 7.3 4.4 5.1 1.7
3.1 2.4 3.7 3.4 4.1 4.4 7.5 7.5 6.1 7.4 5.5 4.2 5.9 6.8 5.1 5.1 1.6
9.6 9.6 9.3 8.7 8.6 5.7 11.5 9.4 11.4 9.4 10.0 12.5 12.2 10.9 10.6 10.0 1.6
10.8 10.3 10.0 8.4 11.6 7.0 11.1 9.0 11.9 9.2 9.9 12.5 14.7 12.5 10.0 10.6 1.8
The measurement of HSV data showed significant improvement in Saturation of the forehead and left cheek and Value of the forehead. In eight
(57%) of 15 Japanese female subjects CAN had whitening effects on the skin of forehead and left cheek with the measurement of the combination
18
A. Kawada et al. / Journal of Dermatological Science 29 (2002) 10–18
of observation and photographs. CAN, therefore, may have whitening effects on skin of the face in the subjects after 6 weeks application. ‘Kusumi’ is a Japanese term that means dullness in color. ‘Kusumi’ collects very much attention from Japanese females as well as whiteness. However, there have been no measurement methods of ‘Kusumi’. Our new method has measured the change in Saturation, suggesting that our method may be used in evaluation of effects of cosmetics on Kusumi.
References [1] Haramoto I, Mizoguchi M. Clinical evaluation of licorice extracts on chloasma. Nishinihon J Dermatol 1995;57:601 – 8 in Japanese. [2] Kita Y, Sugai T. Effect of bleach agents on chloasma. Hifu 1992;34:142 – 6 in Japanese. [3] Sugai T. Recent bleach products seeing from the standpoint of dermatology and scientific evaluation of bleaching effects. Fragrance J 1990;18:32 – 8 in Japanese. [4] Morimoto K, Kawada A, Hiruma M, Ishibashi A. An evaluation of whitening effect of Serum Clarifiant against UVB-induced delayed tanning. Acta Dermatol (Kyoto) 1994;89:675 – 8. [5] Fullerton A, Fischer T, Lahti A, Wilhelm KP, Takiwaki H, Serup J. Guidelines for measurement of skin colour and erythema. A report from the Standardization Group of the European society of contact dermatitis. Contact Dermatitis 1996;35:1 –10. [6] Takiwaki H, Overgaard L, Serup J. Comparison of narrow-band reflectance spectrophotometric and tristimulus colorimetric measurements of skin color. Skin Pharmacol
1994;7:217 – 25. [7] Weatherall IL, Coombs BD. Skin color measurements in terms of CIELAB color space values. J Invest Dermatol 1992;99:468 – 73. [8] Duteil L, Ortonne JP. Colorimetric assessment of the effects of azelaic acid on light-induced skin pigmentation. Photodermatol Photoimmunol Photomed 1992;9:67 – 71. [9] Andreassi L, Flori L. Practical application of cutaneous colorimetry. Clin Dermatol 1995;13:369 – 73. [10] Takiwaki H, Shirai S, Kanno Y, Watanabe Y, Arase S. Quantification of erythema and pigmentation using a videomicroscope and a computer. Br J Dermatol 1994;131:85 – 92. [11] Takiwaki H, Shirai S, Watanabe Y, Nakagawa K, Arase S. A rudimentary system for automatic discrimination among basic skin lesions on the basis of color analysis of video images. J Am Acad Dermatol 1995;32:600 – 4. [12] Smith AR. Color gamut transformation pairs. Comput Graph (ACM) 1978;12:12 – 9. [13] Kawada A, Shiraishi H, Asai M, Aragane Y, Tezuka T. An evaluation of whitening effect of Concentre Antitache Nuit against solar simulated radiation-induced delayed tanning. Skin Res 2001;43:5 – 9. [14] Satoh Y, Kawada A. Action spectrum for melanin pigmentation to ultraviolet light, and Japanese skin typing. In: Fitzpatrick TB, Wick MM, Toda K, editors. Brown Melanoderma. Biology and Disease of Epidermal Pigmentation. Tokyo: University of Tokyo Press, 1986:87 – 95. [15] Buchanan MD, Pendergrass R. Digital image processing: can intensity, hue, and saturation replace red, green, and blue. Electro-Optical Syst Des 1980;12:29 – 36. [16] Joblore GH, Greenberg D. Color spaces for computer graphics. Comput Graph (ACM) 1978;12:20 – 7. [17] Fukue K, Shimoda A, Sakata T. Evaluation of various color models of HIS, in: Proceedings of 1986 Autumn Meeting of Japan Society of Photogrammetry and Remote Sensing, 1986:63 – 68.