Skin analysis


Moisture analysis

Dead skin layer becomes thicker if the amount of moisture on the skin surface is low, and its elasticity and vitality decreased, if not cared properly. Skin also becomes roughen and dark, which are the signs of skin aging.

It is to measure the amount of moisture on the skin surface, contact the moisture sensor to the skin surface and measure the changes of the capacitance of the precision capacitor, which differs depending on the moisture distribution.

Sebum analysis

Sebum is covered the sebaceous gland and connected to the pore. It plays a role as a natural moisturizing cream mixed with sweat in order to prevent excessive loss of moisture. Proper sebum, therefore, is necessary to make facial skin moisturized and to keep skin healthy as it contains mineral water, the nutrition of resident virus who eats extra impurities on our skin.

The sebum tape images and result of detected images:

Sebum smeared in oil paper changes the color of paper to stronger scarlet than the original, therefore, we calculate the value using the color difference of the red areas.

Shininess

The excess oil results from the sebaceous glands in the skin producing the sebum. Those who have this condition usually have shiny skin and large pores.

The captured shine images and result of detected images:

The detected shine region of greater gray intensity value than other regions detected by using the cross mask. The detected region is selected with top 2% brightness pixels.

Pore

The pore is tiny little openings on the layer of the skin where the sebaceous glands are produced by body’s natural oil. Size of the pore may look bigger when; 1) amount of sebum on skin surface secreting from sebaceous glands connected with hair follicle increases, 2) sebum and impurities are piled up inside of pore, or 3) pore wall becomes sagged and stretched by the decrease of elasticity due to skin aging.

The captured pore images and result of detected images:

When it comes to the brightness of the image, pore looks relatively darker than surroundings and has the round shape. First, the dark area is detected, then body hair (similar dark part) is removed from the detected region. Morphology opening method is used in order to separate detected pore which is small, sagged or connected. Since depth and size are important for the value of the pore, the pore index is calculated by multiplying the two values and then dividing it by the size of the image per each pixel.

Wrinkle

Wrinkles are creases, folds, or ridges in the skin. Through exposure to ultraviolet rays, elasticity of skin is impoverished or elastin and collagen are degenerated, which make skin dry and lead to the increase of wrinkle. (Hyaluronan has a strong nature to absorb water and it volumes up to several times if water is kept. On the other hand, however, if water is lost, its bulk decreases with the ratio of square root, cube root, and then wrinkle is created naturally on the skin).

The captured wrinkle images and result of detected images:

Like pore, a wrinkle also looks relatively darker than surrounding areas. So the dark area is first detected and then body hair is removed in a similar way. As the significant components for the wrinkle are length and depth, it is considered as wrinkle only if a length is more than a fixed value. Value of S channel of HLS color space which is similar to real depth is used. For each area finally defined as wrinkle, the length of wrinkle multiplies the average depth value, and the wrinkle index is calculated by dividing it image size value.

Fine wrinkle

Through exposure to ultraviolet rays, elasticity of skin is impoverished or elastin and collagen are degenerated, which make skin dry and lead to the increase of wrinkle.

The captured fine wrinkle images and result of detected images:

A wrinkle also looks relatively darker than surrounding areas. So the dark area is first detected and then body hair is removed in a similar way. As the significant components for the fine wrinkle are length and depth, it is considered as fine wrinkle only if a length is less than a fixed value. For each area finally defined as the wrinkle, the length of wrinkle multiplies the average depth value, and the wrinkle index is calculated by divided it image size value.

Pigmentation

The skin may look darker when melanin pigment is excessively produced or lighter when less produced. This is called “pigmentation” and caused by ultraviolet rays, skin infection or scars.

The captured pigmentation images and result of detected images:

The darker region is detected by brightness value. Through detection process, only the melanin region (brown) is left. Body hair region is also detected, and removed from the melanin region. The numerical value for pigmentation is computed by brightness value. It gives its first weighted value based on the degree of darkness and second weighted value on the degree of darkness against surrounding. The spot index is calculated by combining the two and then dividing it by image size value.

Redness

From a sunburn to an allergic reaction, there are many situations in which your skin can become red or irritated. It may be because extra blood rushes to the skin’s surface to fight off irritants and encourage healing. Skin redness can also come from exertion, such as after a heart-pounding exercise session.

The captured redness images and result of detected images:

The redness region is detected by color difference method. The detected region’s intensity value is different from the surroundings in R and G color channel of RGB image. The redness index is calculated by the difference of two color channels and combination of constant value and then dividing it by image size value.

Porphyrin

Porphyrin is the substance created by acne virus, the main reason for acne. It is indicated as scarlet orange light in response to a specific wavelength range of UV light. The degree of skin disease such as acne can be analyzed by measuring the amount of Porphyrin.

The captured porphyrin images and result of detected images:

There are two types of acne; porphyrin seen as scarlet color and clogged pore seen as halfway between yellow and green color. These can be analyzed separately, but we combine and classify them as acne. We detect porphyrin and clogged pore jointly, and the index is computed by percentage against image size.

Keratin

Skin creates a new cell in every 28 days while pushing out the dead cell. This pushed out cell is commonly called as ‘keratin’. 15-20% of moisture is normal in skin layer but skin becomes drier as the external humidity gets lower due to environment and season. And then keratin occurs in a way to protect skin.

The captured keratin images and result of detected images:

Contact keratin tape to the skin surface, then the amount of dead skin cell on tape is calculated through image analysis algorithm. The detected keratin region on the keratin tape is shown as strong white color. Since lighter white color is sebum, only the value of strong white color is calculated using the difference of brightness.

Sensitivity

Sensitive skin is defined by a few different factors: the thinness of the skin surface, the degree of capillary blood vessels visible on the skin surface, the degree of sensitivity to the season, the external temperature and cosmetic troubles, how often feeling dry etc.

The result process of the sensitive skin analysis:

To diagnose the sensitive skin, we use three different questions that recommended by skin analysis expertise and two different skin diagnosis algorithm. The skin diagnosis algorithm uses the moisture and the redness analysis. The result of the sensitive skin analysis is depending on score of question and diagnosis.

Skin type (Fitzpatrick scale type)

Human skin color ranges in variety from the darkest brown to the lightest hues can be expressed by skin tone and Fitzpatrick scale. The important substance of skin color is the pigment melanin. Melanin is produced in cells called melanocytes, together with the skin, and it is the main determinant of the skin color. Furthermore, the darker skin tends to have larger melanin-making cells which produce more, larger, denser melanosomes, compared to lighter skin.

Two different mode used in skin color detection method (Fitzpatrick scale and Skin Tone):

To detect skin color, the red, blue, and green channel data of skin image is used. The RGB channel data are transferred into CIE L*ab color data by RGB to CIE L*ab calculation method. The L*, a and b pixel values are mapping into color chart data. The color chart data are dependent on skin tone mode or Fitzpatrick scale mode, which we selects in skin color analysis.

Blackhead

Active oil glands under the surface mean large pores at the surface. As excess oil accumulates at the surface, it can combine with shedding skin cells to produce blackheads. Chronic blackheads may stretch pores and make them larger.

The captured blackhead images and result of detected images:

For detecting blackhead, two different images captured by UV and VSL mode are used. First, the dark area is detected from VSL image, and then body hair (similar dark part) is removed from the blackhead candidate region. After first candidate region is detected, the UV image is used to detect the blackhead candidate region, separately. Finally, both UV and VSL candidate region decide the final blackhead region.