Insights · Eyes · Measurement

Palpebral fissure height : length

Vertical eye opening at the pupil divided by horizontal eye length, averaged over both eyes.

Confidence
Tier A · tight interval
Unit
ratio (dimensionless)
Landmarks
up_R, lo_R, up_L, lo_L, en_R, ex_R, en_L, ex_L
CI floor
± 0.03
mean over eyes of dist(upperLid, lowerLid) / dist(en, ex)
Measure yours
Illustration · on your scan the same lines are drawn on your own photo

How to read it

Below the mean

the eyes are relatively narrow vertically for their length ("almond" or hooded configuration). Note this value depends heavily on gaze, fatigue and expression at the moment of capture.

Near the mean

vertical eye opening is typical for the reference group.

Above the mean

the eyes are relatively round/open for their length. As with the low case, this is state-dependent — brow raise and surprise inflate it.

Reference values by population

Reference populationFemale (mean ± SD)Male (mean ± SD)nSource
North American White0.349 ± 0.0430.348 ± 0.0422326Farkas 1994/2005 (NA White young adults) prov.
East Asian0.343 ± 0.0470.344 ± 0.046618Farkas 2005 (Chinese/Japanese/Vietnamese/Thai pooled); Jayaratne 2012 (Southern Chinese, n=103) prov.
South Asian0.349 ± 0.0440.35 ± 0.044210Farkas 2005 (Indian); Metgudmath 2022 prov.
African / African diaspora0.348 ± 0.0430.349 ± 0.043300Farkas 2005 (African-American, Zulu, Angolan) prov.
Middle Eastern0.353 ± 0.0440.354 ± 0.044348Farkas 2005 (Turkish, Iranian, Egyptian); Al-Sebaei 2015 (Saudi, n=168) prov.
Hispanic / Latino0.353 ± 0.0440.353 ± 0.044120Farkas 2005 (Mexican-American) prov.
Not specified — show pooled0.349 ± 0.0470.35 ± 0.0473922Unweighted pool of the six groups above prov.
Where these numbers come from

Transcribed from the cited anthropometric literature; SDs of ratios are propagated conservatively (independence assumed). Entries marked prov. await line-by-line audit against the primary tables — see the norm provenance note.

Why every value has an error bar

Every value Canthion reports carries a 1σ uncertainty propagated from MediaPipe Face Mesh's published accuracy (3.96% of inter-ocular distance mean absolute error per landmark, Kartynnik 2019). For a typical 300 px inter-pupil photo that is ≈12 px per landmark, which is why small ratios and angles get wide intervals. A number without an interval is a number pretending to be more precise than it is.

References
  • Farkas, Katic & Forrest (2005). International Anthropometric Study of Facial Morphology in Various Ethnic Groups/Races. J Craniofac Surg 16(4):615–646. doi.org/10.1097/01.SCS.0000171847.58031.9E
  • Kartynnik, Ablavatski, Grishchenko & Grundmann (2019). Real-time Facial Surface Geometry from Monocular Video on Mobile GPUs. arXiv:1907.06724. arxiv.org/abs/1907.06724

Related: how bone-level measurements are taken from a photo · methodology · all insights