CASZM has released automatic spot measurement. What it addresses is not seeing spots, but turning the color differences in a 2D photograph into 3D data that can be classified, selected and measured by area.
What automatic spot measurement solves
In conventional skin assessment, spot judgement depends heavily on the consultant's experience: on the same face, different people may outline different areas, and at a follow-up visit it is hard to guarantee the same spot is selected. Automatic spot measurement proposes candidate regions by algorithm, then leaves confirmation and correction to the user.
ZETTA's 56-megapixel texture mapping and multispectral imaging give the algorithm enough resolution and spectral information. Red spots, brown spots and UV spots can be detected in separate layers, reducing interference from similar colors.
Six-step detection workflow
Point 01 | Unified multi-spot detection: the intelligent visual perception engine detects red spots, brown spots, UV spots and other skin problems in one pass, across different skin types and lighting conditions.
Point 02 | Skin-tone baseline modeling: multi-level morphological analysis builds a local skin-tone baseline (the native normal skin background), reducing the computational load while preserving detail in high-resolution images.
Point 03 | Candidate region extraction: the algorithm detects suspected spot areas and generates candidate regions, picking up small, dense patches and excluding false spots in non-target areas.
Point 04 | Multi-step fusion and contour refinement: layered results are combined and spot regions merged, with contours smoothed to produce a more natural detection result.
Point 05 | Semantic screening and validation: spot information is read and screened using multi-dimensional features and local contrast to reduce false positives and missed spots.
Point 06 | 3D surface quantification: 2D results are mapped onto a 3D reconstructed facial model to measure true curved surface area, with interactive selection and persistent region highlighting that shows spot area in real time.
From outlining to measuring
The value of automatic spot measurement is not only that it saves the work of manual outlining. Area on a 2D photograph is affected by shooting distance, angle and lens distortion, whereas 3D surface area is closer to the skin's true surface area and better suited to before-and-after comparison.
For clinics this means spot assessment can move from describing an image to recording data: spot count, coverage and area can all enter the report and be repeated with the same definitions at follow-up. The next article covers the features and product advantages of automatic spot measurement.
This article is based on a CASZM WeChat article. View the original article