Moles Results Scheme

Moles are a form of pigmentation of various colors - brown, black, red, purple, and others. They can be located at the level of the skin or raised above it. The outline is often well-defined. The color is saturated, much darker than the overall skin tone. The size varies greatly. Their placement on the face is unpredictable, with no specific patterns of occurrence. Some other forms of hyperpigmentation (post-acne, sunspots) can, as they develop, transition into freckle-like pigmentation.

"algorithm_tech_name": "moles"

Metric

Algorithm returns the amount of moles detected on a face. 

"main_metric": {
                "value": "4",
                "name": "Moles number",
                "tech_name": "moles_number"
            },

Mask

The algorithm returns several vectorised masks of moles for an image aligned with the face.

For a batch mode, all masks for left and right side pictures are attached to the frontal image result.

For single image analysis - only frontal mask is available. For a batch - left, front and right masks are available.

Front Result Mask

  1. "mask_type": "polygon_mask"

  2. "tech_name": "moles_mask_restored"

  3. "side": "front"

  4. features

    • geometry

      • "type": "Multipolygon"

Left Result Mask

"tech_name": "moles_mask_restored"

"side": "left"

Right Result Mask

"tech_name": "moles_mask_restored"

"side": "right"

Example:

Metric structure follows the generic scheme of metric:

Generic Algorithm Metric Scheme

Metric name: pigmentation.moles

Pigmentation PRO App contains masks only for restored images!

JSON object:

Last updated

Was this helpful?