Home › How to find where a photo was taken

How to find where a photo was taken

Five steps, cheapest first — from reading the metadata to reading the picture itself when the metadata is gone.

Somebody sends you a picture and you want to know where it was taken. There is a reliable order to attack this in, and it moves from the cheapest check to the most demanding.

Step 1 — Read the metadata

If the photo came to you as an original camera file, the coordinates may simply be sitting inside it.

iPhone: open the photo, swipe up or tap the ⓘ button. A recorded location shows as a small map.
Android: open it in Google Photos and expand the details panel.
Windows: right-click the file ▸ Properties ▸ Details.
macOS: open in Preview ▸ Tools ▸ Show Inspector ▸ the GPS tab.

If you get coordinates, you are done. Usually you will not: Instagram, Facebook, X, WhatsApp and Discord all strip EXIF on upload, and screenshots never carried it. That is not a bug in your process — it is the normal state of an internet photo, and it is why the remaining steps exist.

Step 2 — Reverse image search

Run the image through Google Lens, Bing Visual Search, Yandex or TinEye. This answers a narrower question than people expect: it tells you whether this picture, or something that looks a lot like it, has already been published somewhere indexed.

That is genuinely useful for landmarks, tourist spots and photos lifted from an article — and useless for an original photo nobody has posted. Reverse search finds copies; it does not infer places.

Step 3 — Read the clues yourself

This is the OSINT core of the job. Zoom in and inventory what the frame actually tells you, in roughly this order of value:

Text. Shop names, street signs, adverts, route numbers, phone numbers. Even unreadable text tells you the script. Readable text can often be searched straight to an address.
Vehicles and plates. Plate colour, shape and proportions identify the country long before the characters are legible. Which side of the road cars drive on halves the world.
Road furniture. Line colours, kerb painting, bollards, guardrails, traffic-light mounting, utility-pole design — all regional standards.
Architecture. Roof pitch, window shape, balconies, materials, render colour.
Vegetation and terrain. Species, undergrowth, soil colour, mountains on the horizon. Often the only signal in a rural photo, and it also dates the season.
Sun and shadows. Shadow direction and length constrain latitude and time of day, and confirm or kill a candidate.

No single clue proves anything. A stack of independent clues that all point the same way does.

Step 4 — Use an AI geolocator

Step 3 is slow, and it is bounded by how much of the world you happen to know. An AI image geolocator does the same reasoning against a far broader sample of the planet, in seconds.

Upload the photo to SpectrAi and you get candidate locations with a confidence score, an error radius, and the clue list behind the answer — so you can check the reasoning rather than take the pin on faith. On a distinctive street it will usually get the city; on a plain interior or an empty beach it will tell you, correctly, that it cannot.

Step 5 — Verify before you believe it

Whatever produced your candidate — you or a model — go and confirm it. Open satellite and street-level imagery of the spot and match something specific: a building corner, a sign, a road layout, the gap between two trees. A prediction is a lead. A matched frame is evidence.

Do not do this to people. Geolocating photos of private individuals in order to find, follow or contact them is harassment, and in many places a crime. Verification, journalism, research and finding your own photos are the point; locating strangers is not.

Frequently asked questions

How can I find out where a photo was taken?

Work through four steps in order: check the EXIF metadata for GPS coordinates; run a reverse image search to see whether the photo or the place is already published; read the visual clues yourself — signage, architecture, plates, vegetation; and run an AI geolocator to generate and rank candidate locations. Then verify the best candidate against street-level imagery.

Can I find where a photo was taken if it came from Instagram or WhatsApp?

Not from metadata — those platforms strip EXIF on upload, so there are no coordinates left in the file. You have to work from the visible content of the image, either by hand or with an AI geolocator.

How do I check a photo’s EXIF data?

On iPhone, open the photo and swipe up or tap the info button; if location was recorded it appears on a small map. On Android, open the image in Google Photos and check the details panel. On Windows, right-click the file, choose Properties and open the Details tab. On macOS, open it in Preview and choose Tools ▸ Show Inspector.

Does reverse image search show where a photo was taken?

Only indirectly. It finds copies or visually similar images that are already indexed, so it helps when the photo has been published before or shows a well-documented landmark. It cannot place a photo nobody has ever posted — that requires content-based geolocation.

Can I find where an old family photo was taken?

Often, yes. Old photos have no metadata, but they are frequently rich in exactly the clues that matter: shopfronts, street furniture, vehicle models, house styles. An AI geolocator can narrow the region, and the vehicles and signage will usually let you date it as well.

Go deeper on the theory in the image geolocation guide, or skip straight to geolocating an image online.

Geolocate a photo with SpectrAi →