Guess My Height

The AI works in two passes. First it hunts for scale: standard building fixtures have standard dimensions, so an interior door at 80 inches or a light switch at roughly 48 inches gives it a ruler it can trust. Then it corrects for the things that fool the eye — a camera held low makes you look taller, one held high makes you look shorter, a phone's wide lens stretches whatever is closest to it, and shoes, thick soles, and hair volume all add apparent inches. What comes back is a range in both feet-and-inches and centimeters, plus which object it scaled you against, so you can judge the reasoning rather than just the number.

A full-body photo taken straight on from about chest height, standing upright and flat-footed, with something of known size in the frame — a doorway, a light switch, a counter, or a staircase works well. Avoid shots taken from below or above, and note that shoes will be counted unless you are barefoot.

Expect a range, not a measurement. With two or three clear reference objects in frame and a straight-on camera, a photo estimate can land close. With none, it is working from body proportions alone — roughly seven and a half head-lengths for an adult — which varies enough between people to be nearly useless on its own, and the tool will say so rather than inventing false precision. Camera angle is the biggest single source of error, and it is invisible in the finished photo: the same person shot from knee height and from eye height can differ by several inches of apparent height. Footwear is the second. If you want a real number, stand against a wall with your heels touching it, mark the top of your head with a flat object, and measure — it takes a minute and beats any photo.

Photographs are terrible at absolute size and always have been. Film and television have exploited this for a century through forced perspective — the technique that made hobbits look small next to wizards without a single special effect, just careful placement of actors at different distances from a single lens. Your brain reads size from context, and when the context is removed or manipulated, it can be fooled completely.

Which is why reference objects do all the real work in a photo height estimate. Building codes and manufacturing conventions mean a surprising number of everyday things are the same height nearly everywhere:

These are US conventions and vary by country and by building age, which is part of why a photo estimate carries a range rather than a figure.

One last thing worth knowing before you compare the result to the number in your head: studies comparing self-reported height against measured height consistently find people round upward, men especially. If the tool guesses shorter than you expect, it is worth checking against a wall before assuming the tool is wrong.