How to Add GPS Location to Scanned Photos
Two ways to add GPS to scanned photos: write known coordinates with your phone or ExifTool, or let AI infer the location from what the image shows.
Scanned prints are the original GPS-blind photograph. When you ran a flatbed scanner over a print from 1968, the scanner wrote its own metadata — scanner model, scan date, color profile — but had no way to know that the shot was taken in front of Niagara Falls or in your grandmother's kitchen in Ypsilanti. The GPS field every photo manager relies on was never filled in because it couldn't be.
Adding it now means one of two things: either you know where the photo was taken and you're writing that knowledge into the file, or you genuinely can't remember and you need help figuring it out. The tools are different, the effort is different, and choosing the wrong path wastes time. This guide covers both.
TL;DR: If you know the location, pull decimal-degree coordinates from Google Maps and write them with your phone's Photos app (free, built in), Apple Photos on Mac (multi-select batch), or ExifTool (command line, scriptable, free). If you don't know the location, AI visual recognition reads the image content — architecture, signs, terrain, vehicles — and proposes a coordinate you review and accept.
Why scanned photos have no GPS
GPS in a digital photo is written by the camera or phone at the moment of capture. A print has no capture moment in the digital sense — the scanner records when you pressed "Scan," not when the shutter tripped in 1972. Flatbed and phone-based scanning apps write their own EXIF (resolution, software, scan timestamp) but leave the GPS field blank because there is nothing to fill it from.
Methods that rely on matching GPS to a GPX track or inferring location from camera metadata won't work here. You are starting from zero.
Path 1 — You know where the photo was taken
If you can identify the place from memory, a caption, or family knowledge, writing the GPS is a five-minute job regardless of how many photos are in the batch.
Get the coordinates first
Open Google Maps, find the exact spot — the street corner, the park entrance, the building — and right-click it (long-press on mobile). The coordinates appear at the top of the context menu; click to copy. You'll get something like 48.8584, 2.2945. That decimal-degrees format is what every editing tool below accepts. The tools handle the internal EXIF conversion to degrees/minutes/seconds; you don't need to compute anything.
Write the GPS
On your phone (one or a few photos)
On an iPhone running iOS 16 or later, open the photo in the Photos app, swipe up to reveal the info panel, and tap Adjust next to the location row (it says "Add Location" if there's none yet). Search for the place by name or drop a pin on a map. The GPS is written directly into the file.
On Android, open the photo in Google Photos, tap the three-dot menu, and choose Edit location. Note that Google Photos sometimes stores the correction in your Google account rather than re-encoding the JPEG itself — if you need the tag physically in the file, export the photo afterward.
In Apple Photos on Mac (a batch from one location)
Select the photos, press ⌘I to open the Info panel, and use the map at the bottom to set the location. You can assign the same coordinates to an entire multi-photo selection in one step — useful for a roll of slides from a single address or a set of scans from one trip.
ExifTool (a folder, or when you need exact coordinates)
ExifTool is the reference command-line metadata editor — free, open source, and scriptable. For a single file:
exiftool -GPSLatitude=48.8584 -GPSLatitudeRef=N \
-GPSLongitude=2.2945 -GPSLongitudeRef=E photo.jpg
To apply the same coordinates to every JPEG in a folder:
exiftool -GPSLatitude=48.8584 -GPSLatitudeRef=N \
-GPSLongitude=2.2945 -GPSLongitudeRef=E /path/to/folder/*.jpg
ExifTool also accepts a CSV mapping filenames to coordinates, so if you have a spreadsheet of scans and their known places, you can tag hundreds of photos in one run. The trade-off is comfort with a terminal.
Path 2 — You don't know where the photo was taken
This is the harder case, and the one traditional tools can't help with. No app can drop a pin when you don't know the city, let alone the street. You need something that works from the image content itself, not from your memory.
AI visual recognition is that tool. A vision model trained on images of places can infer location from cues a person would also use:
- Architecture — roof pitch, window proportions, brick vs stucco vs wood cladding, street layout. A mid-century Levittown ranch house looks nothing like a Spanish-revival bungalow in Pasadena or a brick row house in South Philadelphia.
- Signage — street signs, shop signs, postal codes, license plates. Even partial text narrows a city or country fast.
- Landmarks — cathedrals, bridges, distinctive towers, mountain silhouettes. The easy wins; the model nails them at street-level accuracy.
- Vehicles — a Plymouth Fury in the driveway suggests 1958–1964 North America; a Citroën DS suggests France or French-influenced markets.
- Vegetation and terrain — saguaro cactus means Arizona or Sonora; bare birch trees with snow suggest northern Europe or Canada.
RetroTagr runs this analysis on uploaded scans. Upload the batch, and the AI returns a suggested location with a confidence score for each photo. Accept the high-confidence results in bulk, nudge the medium-confidence ones on a map, and skip or manually tag the photos it can't place — interior shots, tight portraits, featureless suburban scenes.
The free tier covers 100 photos and 5 AI suggestions, which is enough to run a complete first album before deciding whether to go further.
Which method, which situation
| Situation | Best method |
|---|---|
| One or two scans, location remembered, phone in hand | iPhone Photos "Adjust" or Google Photos "Edit location" |
| A batch from one known location, on a Mac | Apple Photos — multi-select, then set location from Info panel |
| Exact coordinates, or a folder with a locations spreadsheet | ExifTool command line |
| Location unknown — can't place from memory or family | AI visual recognition — RetroTagr |
| Mix of known and unknown in one folder | Both paths: manual map for the ones you know, AI for the rest |
Most real scanned albums end up in the last row. A folder of 200 prints typically has some you can place immediately (a famous landmark, a family home you recognize), some you're uncertain about, and some that are genuinely unidentifiable from the image alone. Handling the known ones first with a free tool, then running AI on the remainder, is the most efficient workflow.
Verify the GPS landed
After writing, confirm the tags are present before discarding working copies. Reopen the photo's info panel in Apple Photos (⌘I) or Google Photos — the map preview should show the correct place. From the command line: exiftool -GPS:all photo.jpg prints all GPS fields; both latitude and longitude should appear.
ExifTool creates a backup of the original file with _original appended to the name. Once you've verified the tag is correct, delete the backup with exiftool -delete_original photo.jpg or let them accumulate until you're confident.
Once the GPS is in the file, any app that reads standard EXIF picks it up automatically — Apple Photos puts it on the Places map, Lightroom plots it in the Map module, Google Photos surfaces it in location search, and family-tree software like MyHeritage, RootsMagic, or Ancestry treats it as if the location had always been there.