Skip to content
The Face Shape Detector
Face features

Nose Shape Detector

Measure your nasal index, bridge profile and tip projection from a photo to place you among the eight nose types.

Loading tool…
  • 468

    Landmarks mapped

  • 0

    Data collected

  • Free

    Always, no account

  • Any

    Device or browser

3 simple steps

How the nose shape detector

What happens from the moment you start, in order.

  1. 1

    Frame the nasal landmarks

    With the camera at eye level, MediaPipe locates the nasion between your eyes, the tip-defining points, the subnasale above the lip and both alar rims. Every step runs inside your browser.

  2. 2

    Calculate the nasal index

    Alar width divides by nasal height from nasion to subnasale, giving an index percentage that separates narrow elongated noses from broader ones relative to length.

  3. 3

    Read the profile curve

    A side photo adds dorsal deviation, the offset between your bridge line and a straight nasion-to-tip reference, plus columella angle and tip projection to pin the final type.

What you get

Built to be checked

Every reading comes with the measurement behind it.

Index and angles reported

Your result carries the nasal index percentage, nasolabial angle in degrees and tip projection ratio, so the reasoning behind an assigned type stays visible.

Front and profile analysis

Width and tip definition come from the frontal frame while a dorsal hump or scoop is measured in profile, because no single viewpoint captures both.

Processed in your browser

Google MediaPipe runs locally on your device, so no photo is uploaded, stored or transmitted. The image is gone the moment you close the tab.

Contouring and frame tips

Each type comes with highlight and shadow placement plus eyewear bridge guidance, since this feature reads through light far more than through outline.

Nose type classification (nasal morphology description) names eight combinations of bridge line, tip form and base width: Greek, Roman, snub, nubian, hawk, bulbous, celestial and fleshy. Each name is shorthand for a set of surface structures.

Five measured values decide which name applies. The nasal index divides maximum nostril-to-nostril width by nasal height from nasion to subnasale, and a percentage under about 70 reads narrow while one above about 85 reads broad. Dorsal profile deviation signs the gap between the real bridge line and a straight nasion-to-tip reference, positive for a hump and negative for a scoop. Nasolabial angle runs roughly 90 to 95 degrees in men and 95 to 105 degrees in women. Tip projection, scaled against nasal length, has a conventional reference near 0.55 to 0.60, and tip definition rates how sharply the tip-defining points stand out from the lobule.

Google MediaPipe Face Mesh locates the nasion, the tip-defining points, the subnasale and both alar rims from its 468-landmark topology, running inside your browser. No photograph is uploaded, received or stored at any stage.

What follows covers the troubled history behind nose typing, the landmarks and derived values, the eight types compared in one table, contouring and eyewear guidance, why this feature needs two photographs, and where the readings stop being dependable.

The history worth stating plainly

History matters here more than for any other facial measurement. Nineteenth-century anthropometry used nasal proportions to sort populations, and that work was later put to explicitly racist ends. The descriptive words survived. The interpretive framework attached to them did not, and should not.

What stayed useful is structural. Surgeons and illustrators need terms for the dorsal line, the tip, the base and the nostril aperture, because those features govern how light falls across the center of a face. The eight names carry no information about ancestry, character or health, and this page treats them that way. They rest entirely on the measurements described next.

Landmarks and derived values

Landmarks sit along the bridge, at the nasion, at the tip, at the subnasale and around each alar rim. The five values built from them split cleanly by viewpoint.

  • Nasal index and alar width need the front view only, as plain horizontal and vertical distances.
  • Dorsal deviation needs the profile, taken as a perpendicular offset from the reference line.
  • Nasolabial angle needs the profile, measured between the columella and the upper lip plane.
  • Tip projection needs the profile, scaled as forward extension against nasal length.
  • Tip definition needs the frontal frame, rated by sharpness against the surrounding lobule.

Dorsal deviation is the most informative, which is why a profile shot is not optional.

The eight types compared

TypeDorsal lineTip signature
Greek (straight)Straight, near-zero deviationRefined, moderately projected
Roman (aquiline)Convex, hump in the upper thirdSlightly downward-pointing
Snub (button)Short, slightly concaveSmall, rounded, low projection
NubianLong and mostly straightWide base, broad alar spread
HawkStrongly convex, prominent bridgeSharply downturned, narrow
BulbousVariableWide rounded lobule, poor definition
Celestial (upturned)Slight scoop through the mid-dorsumSmall tip lifted clearly
FleshyStraight to slightly convexThick soft tissue, full nostrils

Columella angles group the eight into five bands. Greek, nubian, bulbous and fleshy all sit at 90 to 100°. Roman runs 85 to 95°, snub 100 to 110°, celestial above 105°, and hawk under 85°.

Greek is the reference straight profile, a continuous line from brow to tip. Roman adds convexity at the bridge without the downturn of a hawk, which is the same form carried much further. Snub and celestial are both short and lifted, but snub is defined by its small rounded tip and celestial by the scooped dorsum leading into the lift. Nubian runs long over a wide base. Bulbous describes the tip alone and sits on any dorsal line. Fleshy describes soft-tissue thickness rather than skeleton, and it turns up more often than the rest. How each reads in a mirror depends on light as much as outline.

Contouring, grooming and frame choice

Light governs nose appearance more than outline does, since this is the most three-dimensional feature on a face and reads through highlight and shadow.

  • Greek noses need almost nothing: one soft highlight down the middle keeps the line clean.
  • Roman noses flatten optically under shadow on the hump with highlight above and below.
  • Snub noses gain apparent length from a narrow full-length highlight, with shadow kept off the tip.
  • Nubian noses narrow when the outer alar edges are shaded and the central highlight stays thin.
  • Hawk noses soften against a fuller textured hairstyle and a defined brow.
  • Bulbous noses gain definition from a small triangle of shadow either side of the tip.
  • Celestial noses exaggerate their lift under tip highlight, so emphasis belongs on the mid-bridge.
  • Fleshy noses control shine under matte products, which otherwise broadens the feature.

Frames follow the same logic. A low or wide bridge pairs with a keyhole or higher-set bridge on glasses, while a prominent dorsum suits frames whose bridge sits below the hump. Both judgments depend on capturing the structure correctly.

Why this feature needs two photographs

Photographs of the front alone cannot finish the job. Width, alar spread and tip definition are frontal quantities. The bridge curve, the columella angle and forward projection exist only in profile.

Shoot the frontal frame at exact eye level, head straight, because dropping your chin foreshortens the feature and inflates the index badly. Then shoot a true side view at ninety degrees, ear visible, gaze horizontal. Light both frames from the front rather than overhead: top light drops a shadow along one side of the dorsum that the mesh can misread as curvature.

Stand at least an arm's length back. Wide-angle front cameras enlarge whatever is closest to the sensor, and on a straight-on selfie that is the feature being measured. This factor explains most complaints that a photo looks unlike the mirror.

Where the readings stop being dependable

Readings from the frontal frame hold up well. Index and alar width are simple distances that repeat closely across photographs taken under matched conditions. Profile readings hold up too, but only at a genuine ninety degrees. Fifteen degrees of rotation flattens an apparent hump and moves the columella angle several degrees.

Type names are the softest part of any result. They cover overlapping regions of a continuous space, and most noses land between two. Fleshy and bulbous describe soft tissue, which the mesh estimates from surface curvature rather than measuring.

This is surface geometry and nothing else. It cannot assess breathing, septal position or sinus function, and no output here is a reason to change anything. Anyone considering surgery may find the vocabulary useful in a consultation, but the assessment belongs to a qualified surgeon with proper imaging.

Eight names, five values, two photographs. The vocabulary survives from anthropometry with its interpretive baggage stripped off, the landmarks split between frontal and profile views, and the table sets each dorsal line, tip signature and columella angle side by side. Contouring works through highlight and shadow, capture demands a level front frame plus a true ninety-degree side frame, and the type names stay softer than the numbers behind them. Nothing here is clinical. Load a pair of photos into the nose shape detector and read the index and the angles, not just the label.

Questions

Nose Shape Detector FAQ

The questions people ask most often about this tool.

What are the main nose types?

Eight descriptive types are in common use: Greek or straight, Roman or aquiline, snub or button, nubian, hawk, bulbous, celestial or upturned, and fleshy. Three structures define them. Whether the bridge line runs straight, convex or concave. How far the tip projects and in which direction it points. And how wide the base sits relative to nasal height.

What is the nasal index?

The nasal index is maximum nostril-to-nostril width divided by nasal height from the nasion to the base of the columella, expressed as a percentage. Values below roughly 70 describe a narrow nose relative to its length, and values above roughly 85 describe a broad one. It is a proportion only, and it carries no meaning beyond describing shape.

Does this tool keep my picture?

No. The nasal landmark mesh is generated by Google MediaPipe executing in your own browser, so the photo never leaves your device. It is not uploaded, not written to any server or database, and not retained after the session. No account is required, and there is nothing for us to delete afterwards.

Can a nose shape result tell me anything medical?

No. This is a surface geometry measurement and nothing else. It says nothing about breathing, septal position, sinus function or any other clinical matter, and it is not a surgical assessment. If you have a health concern or are weighing a procedure, the measurements may give you useful vocabulary, but the evaluation belongs to a qualified clinician.

Which view matters more, the front or the profile?

The profile carries more information. Dorsal deviation, columella angle and tip projection exist only in a side view, and dorsal deviation is the single most informative value in the whole set. The frontal frame contributes alar width, the index and tip definition. Both are needed for a full reading, but a profile alone resolves more types than a front view alone.

Why does my nose look bigger in selfies?

Front-facing phone cameras use short focal length wide-angle lenses that enlarge whatever sits closest to the sensor. On a straight-on selfie taken at arm's reach or nearer, that is your nose, which can appear up to about thirty percent wider than it does in person. Step back further and crop in, or use the rear camera instead.

Start with your face shape

Most styling decisions follow from the shape of your face. Run the detector, then work outward from the result.